The Impact of AI-Driven Recommendation Systems on Patients’ Value Co-Creation Behavior in Online Health Communities: A Moderated Mediation Model

The Impact of AI-Driven Recommendation Systems on Patients’ Value Co-Creation Behavior in Online Health Communities: A Moderated Mediation Model | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The Impact of AI-Driven Recommendation Systems on Patients’ Value Co-Creation Behavior in Online Health Communities: A Moderated Mediation Model Sadia Shabbir, Qiu Rouzhen, Naseeb Khan, Muhammad Usman Arshad This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8413621/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study explores how AI-driven recommendation systems (ADRS) influence patients' participation in value cocreation behavior (VCB) within online health communities (OHCs), aiming to explore the mechanisms behind the fostering of sustainable OHCs. Using a moderated mediation model grounded in social assistance theories and value cocreation, survey data from 450 respondents were analysed via partial least squares structural equation modelling (PLS-SEM) and a multilayer perceptron (MLP) neural network model. The findings reveal that social assistance perceived (SAP) from ADRS positively influences VCB, both indirectly and directly through patient knowledge acquisition (PKA), and that the indirect effect is more robust when patient ability/readiness (PAR) is high. The MLP analysis further confirms the robustness of the model and the substantial moderating role of PAR. This research introduces a novel framework that sheds light on the role of the ADRS in enhancing SAP and PKA, ultimately driving VCB in OHCs. The findings provide practical insights for developing user-centric interventions to foster effective learning and collaboration in AI-driven health platforms, ensuring long-term sustainability and engagement in OHCs. Marketing AI-Driven Recommendation Systems Online Health Communities Value Co-Creation Behavior Patient Knowledge Acquisition Patient Ability/Readiness Social Assistance Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction The rapid evolution of information technology has significantly transformed healthcare delivery, particularly through the growing use of online health communities (OHCs), which have become pivotal in modern healthcare systems. In the modern Era, with fast pacing life, where people require quick answers to their queries, AI-based recommenders play a vital role in fetching the most relevant search or comments while searching online communities. In the United States, recent survey data show that approximately 79% of adults visit online to look up answers about health symptoms or conditions, and approximately 71% use search engines such as Google or Bing at least occasionally to find health information(The Annenberg Public Policy Center of the University of Pennsylvania, 2025 ). According to (Zhao and Liu, 2025 ), 78.6% of people were using the internet in China by December 2024, and approximately 418 million of them had searched for health information and medical services online. The study also highlighted that people frequently use the internet and smartphones to look up health-related information across South Asian countries, including Vietnam, Indonesia, India, the Philippines, Singapore, and Japan. OHCs, such as PatientsLikeMe.com (Szeto et al., 2024 ) and Carenity.com(Bahit et al., 2025 ), have gained popularity, providing patients with a platform to exchange health-related experiences, ask questions, and receive feedback from peers. These communities offer valuable opportunities for patients to collaborate, share health-specific information, and access crucial social assistance (e.g., emotional and informational aid), thereby enhancing patient self-efficacy and bridging gaps in traditional care delivery (Wei, 2025a ; Zhou et al., 2022 ). However, even with their increasing usage, the success and long-term viability of OHCs are often hindered by a lack of patient engagement and contribution. The critical question remains: how can OHCs sustain patient participation over time? Research has shown that value cocreation behavior (VCB) plays a crucial role in maintaining OHCs (Latif and Wang, 2025 ). Traditionally, patients have often been regarded as passive recipients of healthcare services, a model rooted in a one-way, provider-dominated communication flow that severely limits the patient’s capacity to contribute to the overall value creation process (Peng et al., 2022 ). However, the emergence of OHCs has shifted this dynamic, allowing patients to actively participate in knowledge sharing and providing feedback on their health experiences. These interactions, particularly patient-to-patient and patient-to-physician interactions, have been identified as key drivers of VCB (Latif and Wang, 2025 ; Osei-Frimpong et al., 2018 ). Recent advancements in artificial intelligence (AI) have also spurred the adoption of AI-driven recommendation systems (ADRS), including relevant comment fetchers, point search results and social assistance search bots (SSBs), in healthcare settings (Bagheri et al., 2025 ; Gao et al., 2024 ; Grassini et al., 2025 ). These AI tools are increasingly used to provide medical guidance, answer patient queries, and support health management (Al Kuwaiti et al., 2023 ) and have demonstrated their potential during crises such as the COVID-19 pandemic, where recommender systems have played a significant role in reducing the workload of healthcare professionals (Miner et al., 2020 ). In addition to providing technical assistance, AI-based recommenders also provide emotional and informational support, often fostering empathetically and collaborative practices (Sharma et al., 2023 ). This shift toward AI-driven support systems in OHCs highlights the importance of understanding how these tools influence patient behavior in the cocreation process. While prior research has explored the role of human-based social assistance in encouraging VCB (Latif and Wang, 2025 ), there remains a gap in understanding how AI-powered tools such as ADRS can increase patient participation and engagement. Despite the growing use of recommenders and search bots, the mechanisms through which these AI tools influence patients’ collaborative behavior in OHCs are still not well understood. This study aims to fill this gap by examining the role of ADRS in driving VCB within OHCs. Drawing on service-dominant logic (SDL) and value cocreation theory (VCT), this study explores how AI tools such as recommenders and search bots act as resource integrators that encourage VCB in these communities. Moreover, it examines the mediating role of patient knowledge acquisition (PKA) and investigates how patient ability/readiness (PAR) moderates the relationship between social assistance perceived (SAP) from ADRS and VCB. The findings of this study will contribute to the growing body of research on AI in healthcare by providing insights into the role of ADRS in enhancing patient engagement, improving health outcomes, and fostering sustainable OHCs. This study also responds to recent calls for empirical research on how artificial intelligence enables value cocreation in healthcare and digital health ecosystems (Roppelt et al., 2025 ; Swan et al., 2024 ; Xiao and Han, 2025 ) and offers actionable implications for healthcare managers seeking to strengthen patient participation and value cocreation behaviors in online health communities (Liu et al., 2020 ; Wang et al., 2023 ). The study makes three main contributions. Theoretically, it enriches the understanding of AI-enabled value cocreation by integrating trust, learning, and emotional readiness constructs within a single conceptual model. Methodologically, it introduces a hybrid PLS-SEM–MLP framework that merges the strengths of structural modelling and deep learning-based prediction. Practically, it offers insights for AI developers and healthcare institutions in designing systems that foster trust, empathy, and privacy assurance, thereby enhancing patient participation and digital health literacy. Through this integrated perspective, this study aims to advance both the scholarly understanding and practical implementation of AI-driven engagement in healthcare communities. The remainder of this paper is structured as follows. Section 2 develops the theoretical framework and formulates the research hypotheses that form the foundation of the study. Section 3 describes the research framework. Section 4 describes the methodology, including the data collection procedures, sampling details, and measurement instruments. Section 5 presents the data analysis methods, validation tests via PLS‒SEM, and empirical findings, including hypothesis testing, mediation, and moderation analyses. Section 6 details the robustness checks conducted through multilayer perceptron (MLP) modelling. Section 7 discusses the results in light of the literature, outlining the theoretical and practical implications. Section 8 concludes the study by summarizing key findings, acknowledging limitations, and suggesting directions for future research. Collectively, these sections provide a comprehensive understanding of how AI-driven recommendation systems (ADRS) enhance value cocreation behavior (VCB) within online health communities (OHCs) through patient knowledge acquisition and readiness. 2. Theoretical Framework and Hypothesis Development 2.1. Value Cocreation in Online Health Communities (OHCs) Value cocreation (VCC) in online health communities (OHCs) refers to the dynamic process in which patients collaboratively generate mutual value by sharing knowledge, providing emotional support, and actively participating in peer interactions and technology-mediated engagements. Rather than serving as passive recipients of care or information, patients in OHCs take on active roles as coproducers of value, shaping both the content and the quality of community experiences(Ahuja and Shree, 2025 ). Recent research has suggested that AI-based recommendation systems and social assistance search bots can facilitate such cocreation by providing personalized guidance and information, enhancing patients’ self-efficacy and health literacy (Chen & Zhang, 2024). Through interaction with AI, patients learn to interpret health information, share insights, and collaborate with others, thereby improving the collective community value. Accordingly, we propose that AI systems such as search bots and recommendation systems can enhance patient knowledge acquisition and subsequently increase value cocreation behavior (VCB). The results of prior studies regarding VCB in online communities are presented in Table 1. Table I. Prior studies on VCB for online communities Reference (Year) Purpose of Study Types of VCB/Behavior Theories, Sample & Methodology Key Integrated Variables The impact of value co-creation on consumer citizenship behavior: Based on consumer perspective (Yang et al., 2023 ) To examine VCB (value cocreation) in short-video/livestreaming platform context, and how engagement affects citizenship behavior Consumer engagement behavior (participation) & Citizenship behavior Structural Equation Modelling (survey of users of a short-video platform) Consumer engagement, perceived value, citizenship behavior Antecedents and consequences of value co-creation in online shopping (Sanaji et al., 2022 ) To assess how online shopping experience influences VCB and how VCB influences repurchase intention/eWOM Participation behaviour (CPB) & Citizenship behaviour (CCB) PLS-SEM on 195 online-shopping customers (questionnaire survey) Online shopping experience, CPB, CCB, repurchase intention, negative eWOM A Study on Factors Affecting the Value Co-Creation Behavior in Sharing Economy Context (Zou and Shao, 2022 ) To identify factors influencing customer VCB in sharing-economy platform users (e.g. Airbnb) Value co-creation behavior (engagement/citizenship) Survey of 587 sharing-economy platform users; structural modelling (SPSS/AMOS) Corporate (platform) authenticity, corporate image, ethical management; user characteristics; social capital; self-monitoring How is brand love shaped through virtual event value co-creation and sponsorship: a multi-method approach (Xiao Fei et al., 2025 ) To explore how participation and citizenship behaviours in virtual events affect “brand love”, includes VCB in virtual/online event settings Participation & Citizenship behaviour (as components of VCB) Multi-method (survey + qualitative) study analysing virtual event participants. Virtual event co-creation behaviours, sponsorship interactions, brand love/emotional attachment How does the online innovation community climate affect the formation of users’ value co-creation behavior (Tan et al., 2024 ) To examine how community climate (supportive vs. controlling) influences user VCB in online innovation community Participation behavior & Citizenship behavior (user VCB) Empirical analysis using data from a large online community (14 product-sections), behavioural data + modelling (Mplus) Community climate (supportive, controlling), motivation (needs for achievement, power, affiliation), community trust, CPB, CCB Customer citizenship behavior and customer perceived value in China: the mediating role of value co-creation experience (Yin et al., 2025 ) To investigate how Customer Citizenship Behavior (CCB) leads to co-creation experience and perceived value in virtual brand communities Citizenship behavior (CCB) + cocreation experience as mediating process Two-round data collection with total 642 matched questionnaires in virtual brand communities; structural model testing. CCB dimensions, co-creation experience dimensions, customer perceived value (CPV) 2.2. Recommendation Systems and Search Bots in Online Communities Recommendation systems (RSs) and search bots are integral tools for enhancing the user experience within online communities by improving content discoverability and providing personalized recommendations. Recommendation systems optimize the presentation of content by using algorithms to suggest relevant posts, comments, or threads to users on the basis of their interests, preferences, and prior interactions. These systems are particularly useful in communities where users engage with large volumes of content, as they help prioritize the most relevant or helpful information. For example, in health-related forums, when users inquire about specific diseases, recommendation systems can surface the most pertinent responses, expert opinions, or similar discussions, enabling users to access information quickly and efficiently (Jannach and Adomavicius, 2016 ; Ricci et al., 2022 ). Search bots, on the other hand, enhance the traditional search experience by interpreting user queries more effectively, showing the most relevant comments, threads, or posts at the top of search results. These bots often incorporate natural language processing (NLP) and machine learning techniques to better understand and process user input (Qiu et al., 2021 ), ensuring more accurate results. In the context of online communities, search bots can answer user queries instantly by retrieving and ranking content on the basis of various factors, such as textual relevance, user feedback (likes or upvotes), and engagement (Aggarwal, 2015 ; “(PDF) Collaborative Filtering Recommender Systems”, n.d.). The synergy between recommendation systems and search bots ensures that users not only find the right content but are also guided toward further exploration of related topics, fostering a richer, more interactive community experience(Jannach and Adomavicius, 2016 ; Leskovec et al., 2020 ; Qiu et al., 2021 ; Ricci et al., 2022 ). 2.3. Trust in AI Trust in AI refers to users’ belief in the competence, integrity, and benevolence of AI systems (Glikson and Woolley, 2020 ). In the healthcare context, trust determines patients’ willingness to rely on AI recommendations, affecting both acceptance and engagement (Febri Ramanda et al., 2025 ). When patients perceive an AI-driven system as transparent, consistent, and accurate, they are more likely to learn from its feedback and participate actively in online interactions. Conversely, low trust may inhibit knowledge exchange or sharing behaviors. 3. Research Framework and Hypotheses 3.1. Perceived Social Assistance from AI-Driven Recommendation Systems (ADRS) and Value Co-Creation Behavior (VCB) AI-driven recommendation systems (ADRSs) transform patient engagement in online health communities (OHCs) by offering personalized, algorithmically tailored support on the basis of user behavior, preferences, and health data. These systems enhance value cocreation behavior (VCB), a framework in which patients become active contributors to healthcare discussions and decision-making, by fulfilling social and informational needs that are central to social assistance theory (SAT) (Vickery et al., 2025 ). Informational support refers to the delivery of personalized, health-related content, such as treatment options, lifestyle suggestions, and self-care guides, matching patient-specific needs and preferences. Recent studies have demonstrated that such AI-generated recommendations not only increase knowledge but also stimulate information-sharing behaviors (ISB) among patients in OHCs (Yang et al., 2025 ). Informational support plays a pivotal role in empowering users to make informed decisions, increasing their confidence in discussing health challenges within digital communities (Wei, 2025a ). In addition to facts and advice, ADRS increasingly incorporates emotional artificial intelligence (EAI) to simulate empathy through tone-sensitive messaging, sentiment-based prompts, and contextual reassurance. This fosters emotional connection and a sense of psychological safety, which encourages community citizenship behaviors such as feedback and support to peers (Miyazaki and Haderlie, 2025a ). Emotional support is a major predictor of sustained engagement in mental health-focused OHCs, especially among young adults seeking anonymity and empathy (Naga et al., 2025 ). Companionship in ADRS is cultivated through continuous presence, community connectivity features, and AI-enabled nudges that recommend peer groups, shared experiences, and relevant conversations. These interactions reduce patient isolation, foster mutual encouragement, and promote a sustained sense of belonging, which strengthens participation in OHC dialogues(Latif et al., 2024 ). By integrating IS, ES, and COMP, the ADRS drives both participation behaviors (e.g., information sharing, active browsing) and citizenship behaviors (e.g., peer support, feedback provision). These systems act as value enablers within OHCs, enhancing not only individual health literacy but also the collective intelligence and emotional resilience of the patient community. This layered support helps sustain engagement, reduces attrition, and supports long-term digital health transformation (Zhang and Lu, 2025 ). Therefore, the study hypothesizes the following: Y1: Social assistance perceived (SAP) from AI-driven recommendation systems (ADRS) positively impacts patients' value cocreation behavior (VCB) in online health communities (OHCs). 3.2. Mediating Influence of Patient Knowledge Acquisition In recent years, AI-driven recommendation systems (ADRS) have increasingly become foundational tools across online platforms, delivering not only personalized content but also critical social and emotional support to users. In OHCs and other digital environments, ADRS act as adaptive learning companions, enhancing user experiences by dynamically responding to individual information needs (Olawade et al., 2024 ). Through informational support (IS), the ADRS provides customized, health-relevant content, such as lifestyle tips, symptom explanations, and medication guidance, thereby strengthening patients’ knowledge acquisition capacity and autonomy. Recent studies in digital education contexts also show how AI agents significantly enrich learning by delivering real-time, relevant feedback and curated resources tailored to user needs (Vanderhout et al., 2025 ). These systems operate continuously, enabling users to access health knowledge or ask questions at any time. Thus, fostering ongoing, autonomous learning (Miyazaki and Haderlie, 2025a ). Moreover, the emotional support (ES) offered by ADRS plays a key role in reinforcing users’ commitment to learning. Empathetic responses, motivational cues, and contextual affirmations help maintain patients' focus and perseverance, particularly in long-term health management scenarios (Naga et al., 2025 ). ADRS further serve as personalized companions by adapting their guidance to fit each user’s behavioral and emotional patterns, creating engaging and individualized learning experiences (Latif et al., 2024 ). In OHCs, this facilitates the delivery of digestible and conversational health information, empowering patients to engage in interactive, intuitive learning and improving their capacity for health-related decision-making. Y2a: Social assistance perceived (SAP) from AI-driven recommendation systems (ADRS) has a positive effect on patient knowledge acquisition (PKA) in OHCs. Research on digital communities emphasizes the importance of customer learning, where individuals develop expertise through content consumption, discussions, and collaboration. This learning empowers users to become knowledge integrators who contribute value back into the community through engagement and cocreation behaviors(Vickery et al., 2025 ). In OHCs, patient knowledge acquisition (PKA) involves absorbing health knowledge and applying it to personal or shared experiences, which encourages participation in discussions, peer mentoring, and treatment feedback activities (Zhang and Lu, 2025 ). Y2b: Patient knowledge acquisition (PKA) has a constructive effect on patients’ value cocreation behavior (VCB) in OHCs. ADRS streamlines the learning process by minimizing effort and maximizing relevance through real-time interaction and adaptive content filtering. The social interaction mechanisms embedded within ADRS have been shown to improve users' cognitive engagement, enhance collective learning, and drive behavioral participation (Miyazaki and Haderlie, 2025a ). When integrated into OHCs, these features support immersive learning environments, where patients not only acquire knowledge but also engage in collaborative learning loops such as asking questions, offering advice, and sharing experiences with others (Vanderhout et al., 2025 ). This continuous exchange is further strengthened by companionship (COMP) elements of ADRS, which connect patients with similar conditions, reinforcing peer-to-peer learning and emotional bonding (Olawade et al., 2024 ). Emotional reinforcement within these platforms encourages a sense of accountability and prosocial behaviors, such as giving feedback, offering reassurance, or initiating support threads (Latif et al., 2024 ). Despite increasing attention given to PKA in digital health spaces, the specific mechanism by which PKA mediates the link between perceived support and cocreation behaviors remains underexplored. Y2c: Patient knowledge acquisition (PKA) mediates the relationship between social assistance perceived (SAP) from the ADRS and patients’ value cocreation behavior (VCB) in OHCs. 3.3. The Moderating Role of Patient Ability/Readiness Patient ability/readiness (PAR) refers to an individual’s cognitive, emotional, and technical preparedness to learn and adopt new systems or knowledge. In digital health contexts, it encompasses patients’ confidence and ability to interact with AI systems, understand information, and apply it in meaningful ways (Ramadan et al., 2025 ). In particular, within AI-driven recommendation systems (ADRS), PAR significantly shapes how patients perceive and utilize the support provided. For example, patients with greater readiness are better equipped to engage with health information and navigate complex AI interfaces. Conversely, those with limited readiness may require simplified interfaces, structured guidance, and more emotional reinforcement to derive value from ADRS (Ain et al., 2025 ). Given that the ADRS functions as both an educational tool and a support tool in online health communities (OHCs), the impact of social assistance perceived (SAP) on patient knowledge acquisition (PKA) is expected to vary depending on the patient’s level of readiness. High-PAR individuals are more likely to independently engage with IS and seek out ADRS for assistance, whereas low-PAR users may require motivational or emotional nudges to fully participate (Vinh and Hung, 2025 ). Therefore: Y3: Patient ability/readiness (PAR) positively moderates the relationship between social assistance perceived (SAP) from the ADRS and patient knowledge acquisition (PKA) in OHCs. PAR has been recognized across domains as a predictor of digital tool adoption and behavioral intent. Studies on AI integration suggest that users with low digital confidence often disengage or exhibit avoidance behavior, especially when the systems are perceived as complex or ambiguous (Ain et al., 2025 ). In contrast, users with high readiness demonstrate stronger adoption motivation, learning persistence, and peer collaboration, leading to richer interaction with AI systems (Ramadan et al., 2025 ). In the context of OHCs, the importance of incorporating PAR stems from the varying levels of digital literacy and self-efficacy among patients. Those who possess the ability to interact effectively with ADRS are more likely to leverage SAP for deeper learning and value cocreation behavior (VCB). In contrast, those with lower readiness may encounter cognitive or emotional barriers, reducing their ability to participate in and benefit from community engagement activities (Vinh and Hung, 2025 ). This underscores the need to consider PAR in a moderated mediation framework, where it not only influences the direct relationship between the SAP and PKA but also affects the strength of the indirect path from the SAP to the VCB via the PKA. In such a framework, we anticipate the following: Y4: The indirect effect of SAP from ADRS on patients' VCB through the PKA will be stronger when patient ability/readiness (PAR) is high. As shown in Fig. 1 , this moderated mediation model captures the conditional nature of ADRS efficacy in OHCs. It posits that ADRS can enhance patient outcomes across the board, but their impact will be magnified for those with greater readiness, who are better positioned to engage, learn, and contribute. On the other hand, individuals with lower readiness may experience attenuated benefits, necessitating adaptive system designs that account for digital literacy variability and learning support needs. A summary of the above results is presented in Table 2. Table II. Summary of Constructs and Hypotheses Construct Definition Hypothesis Expected Relationship Social Assistance Perceived (SAP) Support provided by AI-driven recommendation systems (ADRS) Y1 + on PKA Patient Knowledge Acquisition (PKA) The process by which patients acquire knowledge from ADRS and peers Y2 + on VCB Patient Ability/Readiness (PAR) The capacity and readiness of patients to interact with ADRS and engage Y3 Moderates the effect of SAP on PKA Patient Knowledge Acquisition (PKA) Mediator between SAP and VCB in OHCs Y4 Mediates the relationship between SAP and VCB 4. Methodology 4.1 Research Design This study employs a quantitative, cross-sectional approach to empirically evaluate the theoretical framework addressing AI-enabled value cocreation within online health communities (OHCs). Data were collected via a structured online questionnaire administered to patients who are active participants in OHCs that incorporate AI-based tools such as recommendation systems, conversational agents, or personalized digital health assistants. For data analysis, this study uses partial least squares structural equation modelling (PLS-SEM) to validate the proposed model and multilayer perceptron (MLP) techniques to investigate nonlinear relationships and assess the relative importance of predictors. The integration of these methods is in line with current best practices for analysing complex models in technology adoption research (Ketchen, 2013 ). 4.2 Data Collection Data for this study were collected from active members of online health communities (OHCs) in China, with a focus on both patient–to-patient and patient–doctor interactions. To ensure contextual relevance, participants were presented with examples of six well-known Chinese OHC platforms, each hosting between approximately 450,000 and 3.8 million registered members, demonstrating high levels of digital engagement. The study specifically targeted individuals who had experience using AI-driven recommendation systems (ADRS) embedded within these OHCs, as these systems assist users by offering tailored health recommendations, emotional guidance, and informational support. Within these online communities, members frequently engage in collaborative health-related behaviors such as exchanging medical advice, sharing treatment experiences, offering feedback on healthcare services, and supporting other patients’ recovery processes. Given their interactive and participatory nature, OHCs provide an ideal environment for examining value cocreation behavior (VCB). Prior to data collection, ethical approval was granted by the university’s Institutional Ethics Review Committee. The study adhered strictly to ethical research standards by ensuring participant confidentiality, voluntary participation, and informed consent. All respondents were informed that the collected data would be used solely for academic purposes and that anonymity would be preserved. The data for this study was collected using random sampling, a probability sampling technique where participants are selected at random from the population. This approach ensures that every individual has an equal chance of being included in the sample, which helps to reduce bias and increase the generalizability of the results. Two screening questions were used to ensure sample validity. First, participants were asked to identify the OHCs they actively used to verify their familiarity with such platforms. Second, they were asked to confirm whether they had recently interacted with ADRS tools within these communities. Only respondents meeting both criteria were permitted to complete the main questionnaire, which was administered online. The survey, designed to be completed within 10–12 minutes, was distributed over a three-month period from March 2025 to May 2025. Out of the 780 distributed questionnaires, 476 responses were received. After incomplete and inconsistent responses, were excluded, 362 valid responses were retained for the final analysis. The demographic profile of the respondents is presented in Table 3, and our expected data characteristics are presentd in Table 4. Table III. Sample characteristics Demographics Items Frequency Percentage (%) Gender None None Male 200 57.0 Female 162 43.0 Age None None 21–30 131 38.0 31–40 120 34.0 41–50 58 17.0 51–60 42 13.0 61 or above 12 3.5 Schooling Level None None High school or lower 38 8.0 Bachelor’s degree 235 67.5 Master’s degree or higher 95 24.0 Tenure of using OHC None None 37 months 15 2.5 Recent experience with OHC None None < 7 days 75 22.0 7 days to < 15 days 130 38.0 15 days to < 1 month 110 32.5 1 month to < 2 months 40 7.5 Table IV. Expected Data Characteristics Statistic Expected Range Notes Mean age 25–45 years Active digital health users Gender ~ 55% female Common in OHC participation OHC experience 6–24 months Moderate engagement Reliability (α) 0.80–0.94 Indicates high internal consistency R² (VCB) 0.45–0.60 Substantial explanatory power 4.3 Construct Measurement The measures used in this study were adapted from existing scales in prior research to ensure consistency with established methodologies (Liu et al., 2020 ; Yi and Gong, 2013 ). To assess each construct, a 7-point Likert scale was used, where 1 indicated “strongly disagree” and 7 indicated “strongly agree.” The items for each construct were slightly modified to suit the specific context of this study, particularly in relation to online health communities (OHCs) and AI-driven recommendation systems (ADRS). Since the research was conducted in China, the questionnaire followed a double translation protocol (Harkness et al., 2010 ). Initially, the English version was translated into Chinese, and then translated back into English by two bilingual experts from the management faculty, who were familiar with the technical terminology but were unaware of the study's aims. The back-translated versions did not reveal any significant differences from the original versions. The dependent variable for this study, patients' value cocreation behavior (VCB), was measured via items adapted from (Yi and Gong, 2013 ) and (Liu et al., 2020 ). VCB can be divided into two main categories: participation behavior and citizenship behavior. Participation behavior was further subdivided into information sharing behavior (ISB) and responsible behavior (RB), with ISB including 5 items and RB consisting of 4 items. Citizenship behavior was measured via Advocacy Behavior (AB) and Feedback Behavior (FB), with each containing 3 items. The relationships among participation behavior, citizenship behavior, and VCB can be expressed as a weighted sum, where VCB is a function of participation and citizenship behaviors. The formula for VCB is Eq. 1. \(\:VCB=\alpha\:\cdot\:PB+\beta\:\cdot\:CB\) Eq. 1 where PB is the total number of participation behaviors (ISB + RB), CB is the total number of citizenship behaviors (AB + FB), and α and β represent the weights of each behavior type on the basis of their influence on VCB. For the independent variable, social assistance perceived (SAP) from AI-driven recommendation systems (ADRS) in OHCs, the construct was operationalized as a multidimensional variable, following (Liu et al., 2020 ). The SAP is broken down into three key dimensions: informational support (IS), which consists of 3 items; emotional support (ES), which comprises 4 items; and companionship (COMP), which has 2 items. The overall SAP construct was calculated as the sum of these three dimensions, weighted by their respective importance in the context of ADRS. This can be represented as Eq. 2. \(\:PSS=\sum\:({\omega\:}_{i}\cdot\:I{S}_{i})+\sum\:({\theta\:}_{j}\cdot\:E{S}_{j})+\sum\:({\varphi\:}_{k}\cdot\:COM{P}_{k})\) Eq. 2 where ω, θ, and φ represent the weights for each dimension, and IS, ES, and COMP refer to informational support, emotional support, and companionship, respectively. The mediator variable, patient knowledge acquisition (PKA), was adapted from (Zhao, 2025 ) and consists of 3 items tailored for the OHC context. The PKA represents the process through which patients acquire knowledge from interactions within these communities, particularly those mediated by. The relationship between SAP and PKA was modelled via a linear regression approach, where PKA is predicted by the social assistance perceived (SAP): \(\:PKA={\delta\:}_{1}\cdot\:SAP+ϵ\) Eq. 3 where δ1 is the coefficient representing the effect of the SAP on the PKA, and where ε is the error term. For the moderator variable, patient ability/readiness (PAR), the scale was adapted from (Gao et al., 2023 ) with 4 items adjusted for the OHC context. PAR captures the readiness and digital competence of patients to engage with ADRS. The moderating effect of PAR on the relationship between SAP and PKA was modelled via an interaction term, which captures how PAR influences the effect of SAP on PKA. This relationship is expressed as: \(\:PKA={\delta\:}_{2}\cdot\:SAP+\gamma\:\cdot\:PAR+\eta\:\cdot\:(SAP\times\:PAR)+ϵ\) Eq. 4 where δ2, γ, and η are the coefficients for each term, representing the direct effects of SAP, PAR, and the interaction between SAP and PAR, respectively. Finally, the study includes five control variables: gender, age, education, recent experience with OHCs, and tenure of use (Bu et al ., 2022). These control variables were included to account for potential confounding factors, allowing for a more accurate analysis of the primary constructs. The impact of these control variables on VCB was modelled as part of a multiple regression analysis: \(\:VCB={\beta\:}_{1}\cdot\:SAP+{\beta\:}_{2}\cdot\:PKA+{\beta\:}_{3}\cdot\:PAR+\sum\:({\theta\:}_{n}\cdot\:Contro{l}_{n})+ϵ\) Eq. 5 where Control_n represents the control variables (e.g., gender, age), and β1, β2, β3, and θn are the coefficients for each respective term. This model allows the effects of the primary constructs to be isolated while accounting for the influence of demographic and contextual factors. 5. Data Analysis and Validation 5.1 Data Analysis The data were analysed in two phases. First, we used SmartPLS 4.0 to perform partial least squares structural equation modelling (PLS-SEM) (Ringle et al ., 2021). This method has gained significant traction across various disciplines because of its ability to handle complex models and data with nonnormal distributions, which are often encountered in social science research (Sarstedt et al ., 2021). PLS-SEM is especially valuable when relationships between exogenous and endogenous variables are nonlinear. Linear models can fail to capture the intricacies inherent in phenomena such as value cocreation behavior (VCB). To address this issue, we integrated a neural network approach, which is capable of modelling both linear and nonlinear relationships among constructs. Scholars have increasingly advocated the use of hybrid approaches that combine PLS-SEM with machine learning techniques to increase model robustness and generalizability (Hair et al ., 2020). Furthermore, to test the reliability and validity of the study model, we utilized a multilayer perceptron (MLP), following the recommendations of recent advancements in methodology (Shahzad et al ., 2021). This dual-approach methodology provided a more comprehensive understanding of the relationships among the key variables in our model. Table V. Data Analysis Summary Stage Method Purpose Output 1 PLS-SEM(SmartPLS 4) Test hypotheses, mediation, moderation Path coefficients, \(\:{R}^{2}\) , f², VIF 2 MLP (IBM SPSS) Identify variable importance, capture nonlinearity RMSE, R², Normalized Importance 3 Bootstrapping Confirm robustness Confidence Intervals (CI), significance 5.2 Measurement Validation To assess the measurement model and test the hypotheses, we employed SmartPLS 4 (Ringle et al., 2024 ). Following the guidelines of (Latan et al., 2023 ), a series of validity and reliability tests were performed. The internal consistency of the constructs was validated via Cronbach’s alpha (α) and composite reliability (CR), both of which exceeded the recommended threshold of 0.70, indicating that the constructs are both reliable and consistent. Additionally, the average variance extracted (AVE) was calculated, with all values exceeding the minimum threshold of 0.50, confirming adequate convergent validity (Hair and Alamer, 2022 ). The results for α, CR, AVE, and factor loadings are summarized in Table 6. Given that AI-driven recommendation systems (ADRS) and value cocreation behavior (VCB) are modelled as second-order formative constructs, traditional reliability and validity tests are not suitable because of their directional causality and expected lower correlations among formative indicators, which should not be disregarded (Hair et al., 2022 ). As per (Hair et al., 2022 ), we assessed the outer weights for the ADRS and VCB dimensions, ensuring that they met the significance threshold of p < 0.001. Furthermore, multicollinearity tests were conducted for the second-order formative constructs, and the variance inflation factor (VIF) values for the measured items were all below the threshold of 5.0, confirming that multicollinearity issues were not present. These results, presented in Table 7, indicate that the second-order formative constructs exhibit strong validity, and no multicollinearity concerns were identified. Table VI. Reliability and Validity of the Constructs Construct Items Loadings α CR Social Assistance Perceived (SAP) from ADRS 0.918 0.912 0.586 Informational Support (IS) IS1, IS2, IS3 0.885 0.889 0.668 Emotional Support (ES) ES1, ES2, ES3, ES4 0.824 0.899 0.742 Companionship (COMP) COMP1, COMP2 0.795 0.911 0.829 Patient Knowledge Acquisition (PKA) PKA1, PKA2, PKA3 0.764 0.872 0.678 Patient Ability/Readiness (PAR) PAR1, PAR2, PAR3, PAR4 0.836 0.889 0.674 Value Co-Creation Behavior (VCB) 0.935 0.939 0.523 Information Sharing Behavior (ISB) ISB1, ISB2, ISB3, ISB4, ISB5 0.885 0.912 0.682 Responsible Behavior (RB) RB1, RB2, RB3, RB4 0.818 0.889 0.735 Feedback Behavior (FB) FB1, FB2, FB3 0.798 0.883 0.718 Table VII. Evaluation of Second-Order Formative Constructs with VIF Values for Latent Variables Constructs Weight Coefficients T-Statistics P Values VIF (Variance Inflation Factor) Value Co-Creation Behavior (VCB) 0.253 23.92 0.0 2.27 Value Co-Creation Behavior (VCB) 0.229 24.12 0.0 2.375 Value Co-Creation Behavior (VCB) 0.422 30.12 0.0 3.63 Value Co-Creation Behavior (VCB) 0.245 25.31 0.0 2.31 AI-Driven Recommendation Systems (ADRS) 0.505 31.49 0.0 2.235 AI-Driven Recommendation Systems (ADRS) 0.358 24.19 0.0 2.03 AI-Driven Recommendation Systems (ADRS) 0.269 23.5 0.0 1.88 5.3 Common Method Bias (CMB) Common method bias (CMB) remains a concern in survey-based research, particularly when data are collected from a single source (Podsakoff et al., 2024 ). To address this issue, we employed two well-established techniques. First, we conducted Harman’s single-factor test (Kock, 2020 ), which revealed that no single factor explained more than 38% of the variance, well below the conventional threshold of 50%. This finding indicates that CMB is not a significant issue in our study. The second method involved performing a full collinearity assessment via SmartPLS 4, following the approach suggested by (Kock, 2015 ). This technique, which is gaining widespread acceptance in contemporary research, assesses multicollinearity and ensures the validity of the structural model. The variance inflation factor (VIF) values reported in Table 8 indicate that most values are well below the threshold of 3.3. Only one value, ISB (6) (VIF = 3.369), exceeded this threshold, but it remains within the acceptable range of 5, as recommended by (Kock, 2020 ) Both methods confirm that CMB is not a significant concern in our study. Table VIII. Full Collinearity Test of First-Order Constructs on VCB Constructs VIF Advocacy Behavior (AB) → Value Co-Creation Behavior (VCB) 2.342 Companionship (COMP) → Value Co-Creation Behavior (VCB) 2.315 Emotional Support (ES) → Value Co-Creation Behavior (VCB) 2.731 Feedback Behavior (FB) → Value Co-Creation Behavior (VCB) 2.305 Informational Support (IS) → Value Co-Creation Behavior (VCB) 2.149 Information Sharing Behavior (ISB) → Value Co-Creation Behavior (VCB) 3.42 Patient knowledge Acquisition (PKA) → Value Co-Creation Behavior (VCB) 2.89 Patient Ability/Readiness (PAR) → Value Co-Creation Behavior (VCB) 2.426 Responsible Behavior (RB) → Value Co-Creation Behavior (VCB) 2.318 5.4. Discriminant Validity We applied two widely recognized methods to assess discriminant validity. The first approach used to evaluate discriminant validity in this study was Fornell and Larcker’s criterion (1981), which posits that the square roots of the average variance extracted (AVE) for each construct should exceed the correlations between the constructs. The second method applied was the heterotrait-monotrait (HTMT) ratio of correlations, as suggested by (Henseler et al., 2015 ). According to this more recent method, the HTMT value should remain below the threshold of 0.85. The results obtained from both techniques in this study satisfy the necessary criteria. Additional details are provided in Tables 9 and 10. Table IX. Fornell-Larcker Criterion Construct C1 C2 C3 C4 C5 C6 C7 C8 AB 0.867 0.456 0.574 0.648 0.52 0.672 0.68 0.586 COMP 0.466 0.916 0.624 0.682 0.547 0.645 0.722 0.655 ES 0.563 0.665 0.874 0.52 0.609 0.773 0.671 0.698 FB 0.649 0.503 0.521 0.836 0.633 0.596 0.741 0.71 IS 0.494 0.612 0.687 0.546 0.857 0.532 0.7 0.623 ISB 0.675 0.741 0.751 0.689 0.532 0.831 0.691 0.763 PKA 0.634 0.688 0.696 0.607 0.612 0.719 0.834 0.741 PAR 0.625 0.65 0.602 0.558 0.539 0.694 0.732 0.845 RB 0.615 0.562 0.527 0.633 0.563 0.703 0.624 0.849 Table X. Heterotrait-Monotrait Ratio (HTMT) Construct C1 C2 C3 C4 C5 C6 C7 C8 AB 0.88 0.626 0.704 0.749 0.633 0.77 0.688 0.628 COMP 0.578 0.812 0.813 0.659 0.693 0.818 0.785 0.692 ES 0.659 0.79 0.787 0.708 0.69 0.838 0.802 0.77 FB 0.81 0.633 0.653 0.89 0.692 0.782 0.81 0.713 IS 0.58 0.732 0.792 0.741 0.721 0.85 0.819 0.781 ISB 0.758 0.763 0.753 0.8 0.702 0.85 0.812 0.749 PKA 0.758 0.802 0.824 0.797 0.759 0.838 0.819 0.76 PAR 0.705 0.74 0.681 0.658 0.622 0.808 0.831 0.735 RB 0.72 0.686 0.654 0.801 0.694 0.818 0.823 0.761 The relationships between the underlying variables were examined via R² values. According to (Hair et al., 2019 ), R² values of 0.25, 0.50, and 0.75 correspond to weak, moderate, and substantial explanatory power, respectively. In our model, the R² values were 0.552, 0.540, and 0.633, indicating moderate to substantial levels of explanation. Furthermore, the structural model results met the goodness-of-fit criteria, with RMSEA = 0.053 (below the threshold of 0.080) and SRMR = 0.073 (also under the acceptable threshold of 0.080). Thus, all model fit criteria were satisfied (Hair et al., 2019 ). To verify the results, we performed a bootstrap procedure with 5,000 samples to test the significance of all paths in the structural model (Hair and Alamer, 2022 ). The results of the hypothesis tests for this study are shown in Table 11, and they were analysed in three stages. Table XI. Results of Hypothesis Testing Variables Model 1: β Model 1: T-statistics Model 2: β Model 2: T-statistics Model 3: β and T-statistics R² (ADRS) to (VCB) 0.733 9.811 0.433 5.702 0.428 0.552 (ADRS) to Patient Knowledge Acquisition (PKA) 0.822 11.123 0.315 2.512 0.540 Patient Knowledge Acquisition (PKA) to (VCB) 0.385 5.213 0.385 5.213 0.633 (ADRS) → Patient Knowledge Acquisition to VCB 0.315 4.671 (PAR) to (VCB) -0.185 2.492 PAR × ADRS to PKA 0.213 5.023 In Model 1, we initially tested the direct impact of AI-driven recommendation systems (ADRS) on value cocreation behavior (VCB). The results indicated a significant positive relationship between ADRS and VCB (β = 0.733, t = 9.811, p < 0.001), providing strong support for H1. Next, in Model 2, we explored a mediation model based on the framework introduced by (Latif et al., 2024 ), which considers both direct and indirect pathways. The results demonstrated that the ADRS positively influences patient knowledge acquisition (PKA) (β = 0.822, t = 11.123, p < 0.001), and PKA, in turn, significantly impacts VCB (β = 0.385, t = 5.213, p < 0.001). These findings support H2a and H2b. The indirect effect in Model 2 (β = 0.315, t = 4.671, p < 0.001) further confirms the mediation process. The reduction in effect size when PKA is included in the model suggests partial mediation, confirming H2c. In Model 3, we examined the moderating role of Patient Ability/Readiness (PAR). Initially, we tested the direct effect of PAR on the relationship between ADRS and VCB (β = -0.185, t = 2.492, p < 0.05) without considering the interaction effect. Next, we evaluated the interaction effect of ADRS and PAR on VCB, and the results indicated that PAR significantly moderated the relationship between ADRS and PKA (β = 0.213, t = 5.023, p < 0.001), thus supporting H3. The significant interaction slopes are visually represented in Fig. 2 . Finally, we conducted a moderated mediation analysis via bias-corrected bootstrapping, as outlined by Preacher et al . (2023). The results presented in Table 11 show that the indirect effect of the ADRS on VCB through the PKA varies at different PARs. Specifically, when the PAR is high, the indirect effect of ADRS on VCB is significant (β = 0.186, 95% bias-corrected CI [0.103, 0.269]), whereas when the PAR is low, this effect becomes nonsignificant (β = 0.038, boot SE = 0.095, 95% bias-corrected CI [-0.067, 0.136]). Therefore, the positive indirect effect of the ADRS on VCB through the PKA is present when the PAR is high, supporting H4. The control variables in the study, including gender (β = -0.041), age (β = 0.032), education (β = 0.011), recent experience with OHCs (β = 0.015), and tenure of OHC use (β = 0.012), were found to be insignificant. 6. Robustness Checks: Multilayer Perceptron (MLP) Approach To validate the proposed research model and examine the hypotheses, this study employs a hybrid analytical approach that integrates structural equation modelling (SEM) and a multilayer perceptron (MLP). Initially, PLS-SEM was utilized to assess the reliability and validity of the measurement model, with a primary focus on evaluating the constructs and their indicators. Afterward, the structural model was analysed to verify the hypothesized relationships among the constructs. Subsequently, MLP analysis was performed to increase the model's accuracy, further identifying the critical factors that influence value cocreation behavior (VCB). Table XII. Conditional Indirect Effect of the ADRS on VCB, through the PKA at Different PAR Values Boot Effect LLCI ULCI PAR Values + 1 SD 0.192 0.109 0.276 -1 SD 0.047 -0.072 0.15 Mean 0.128 0.039 0.205 Compared with conventional regression methods, multilayer perceptron (MLP) methods are recognized for their superior predictive power, offering more accurate forecasts (Kadir et al., 2025 ; Schroer and Just, 2023 ). Previous studies suggest that integrating structural equation modelling (SEM) with multilayer perceptron (MLP) techniques can significantly advance the field of information systems, potentially establishing a new paradigm for research methodologies (Leong et al., 2025 ). In this study, we applied the widely used multilayer perceptron (MLP) method, a common artificial intelligence approach, to train neural networks (Drałus et al., 2023 ). The neural network model was built via SPSS 21.0, which incorporates multiple hierarchical layers. Typically, an MLP consists of an input layer, one or more hidden layers, and an output layer. While no standardized method exists for determining the optimal values, the configuration of the hidden layers is dependent on the complexity of the problem being addressed (Trigka and Dritsas, 2025 ). The significance of the predictor variables was evaluated in two stages. Initially, three key covariates were introduced in the input layer, the SAP from the ADRS, the PKA, and the PAR, with VCB placed in the output layer. The sigmoid function was used as the activation function in both the hidden and output layers, following the methodology outlined by (Tao et al., 2025 ) As recommended by (Tao et al., 2025 ) and (Drałus et al., 2023 ), the MLP model was validated by testing with hidden nodes ranging from 1–10. To mitigate overfitting, we employed tenfold cross-validation, with 70% of the data used for training and the remaining 30% for testing. The root mean square error (RMSE) was calculated to assess errors during both the training and testing phases, as shown in Table 13. The findings revealed that the average RMSE for value cocreation behavior (VCB) was 0.370 for the training set and 0.363 for the testing set, indicating minimal variation. These low values suggest a high degree of accuracy in predicting VCB, the endogenous construct. The detailed results are presented in Table 13. Furthermore, a sensitivity analysis was performed to assess the contribution and importance of the covariates. The model's importance and normalized importance were evaluated by averaging the generated values of the constructs over ten iterations to predict the output. Normalized importance refers to the proportion of each input relative to the highest value, highlighting that patient ability/readiness (PAR) was the most influential predictor of VCB (importance value = 0.422), followed by SAP from ADRS (0.307) and PKA (see Table 14). This underscores the importance of optimizing the most influential predictor to increase the effectiveness and overall performance of the model. While some minor adjustments in the model ranking were observed, the PAR and SAP from the ADRS maintained similar rankings throughout. Table XIII. Neural network validation for training and testing data Neural Networks Training Data N Training Data RMSE Testing Data N Testing Data RMSE Sum of Square Error 1 230 0.395 108 0.317 37.832 2 230 0.372 108 0.43 33.409 3 234 0.338 104 0.327 25.297 4 247 0.357 91 0.345 32.678 5 227 0.421 111 0.412 41.213 6 237 0.346 101 0.392 26.045 7 238 0.386 100 0.315 36.132 8 242 0.342 96 0.367 29.812 9 237 0.359 101 0.353 29.431 10 244 0.345 94 0.398 28.914 Average 0.370 0.363 SD 0.032 0.039 7. Discussion 7.1. Summary of Key Findings This study aimed to develop and test a moderated mediation model explaining the formation of patients' value cocreation behavior (VCB) in online health communities (OHCs). The results indicated that the support provided by AI recommender systems (AIRS) positively influences patients' VCB, which is consistent with prior research in various ways, despite the relatively underexplored context and the novel nature of our model. Specifically, similar to the findings of (Li et al., 2024 ), social assistance plays a crucial role in influencing individuals' willingness to engage in community-based activities in digital commerce settings. However, our study uniquely focused on the specific contributions of various types of social assistance in shaping patient behavior toward value cocreation. Additionally, our findings align with the results of (Liu et al., 2020 ), who demonstrated that social assistance is multidimensional and affects cocreation intentions. Unlike previous studies, this research incorporates multiple dimensions of social assistance, such as informational support (IS), emotional support (ES), and companionship (COMP), all of which significantly influence patients' VCB. We conceptualized SAP from ADRS as a second-order formative construct, as their integration provides a more holistic understanding of the latent variable. Furthermore, the findings expand upon the work of (Latif et al. , 2025), reinforcing the idea that social assistance is an essential factor in the development of VCB. Table XIV. Sensitivity Analysis Neural Networks SAP from ADRS PAR PKA 1 0.391 0.379 0.24 2 0.118 0.591 0.296 3 0.362 0.421 0.225 4 0.368 0.42 0.22 5 0.359 0.3 0.34 6 0.312 0.502 0.209 7 0.261 0.528 0.218 8 0.455 0.275 0.261 9 0.135 0.51 0.359 10 0.253 0.396 0.348 Average Importance 0.307 0.422 0.271 Relative Importance 0.726 1 0.643 Normalized Importance 72.66 100 64.32 This study highlights the influential role of AI-based social assistance recommender systems (ADRSs) in inducing members' value cocreation behavior (VCB) in online health communities (OHCs), which is a traditionally human-driven process. Our research suggests that ADRS are more effective in fostering patient VCB within OHCs because of their continuous support, which helps patients remain engaged with others. For example, a search bot providing reliable informational support (IS) and companionship (COMP) encourages patients to interact confidently and contribute to community discussions. Recent research by (Li and Tuunanen, 2022 ) explores how social interactions and resource integration contribute to information technology-based value cocreation in service systems. This study further demonstrated that AI tools, such as ADRS, enhance patients' VCB. Unlike (Latif et al. , 2025), who focused on the role of social assistance search bots in addressing global health challenges, our study specifically investigates how various types of social assistance, as perceived by recommenders and search bots, influence patient participation and contribution behaviors in OHCs to improve individual and collective health outcomes. The results of this study align with those of previous studies indicating that learning stems from multiple types of social assistance (Zhao, 2025 ). For example, IS provided by search bots is instrumental in helping members learn in educational communities (Guan et al., 2025 ).Our findings indicate that ADRS functions as an antecedent to learning by providing patients access to relevant information and empathetic responses to their concerns, thereby enhancing their overall experience and engagement. Consistent with (Nohutlu et al., 2023 ), who emphasized that learning is a key motivator for customer value cocreation in online innovation communities, our study shows that patient knowledge acquisition (PKA) positively impacts VCB in OHCs. The positive influence of these variables suggests that patients in OHCs acquire knowledge and specific skills to address their health issues, which ultimately drives feedback behavior (FB). This aligns with the work of (Latif et al. , 2025) and (Chen et al., 2019 ), confirming that learning serves as a mediator to explain how patients use ADRS to generate various VCBs. Unlike earlier studies by (Liu et al., 2020 ) and (Latif et al. , 2025), which examined member belongingness and ethical perception as mediators between social assistance and VCB, our study revealed that, in a supportive learning environment, patients feel more empowered and positive about participating in cocreation activities. Consequently, learning emerges as a crucial component of the sustainability of OHCs (Chen et al., 2019 ). Additionally, our findings show that patient ability/readiness (PAR) positively moderates the relationship between PKA and ADRS in OHCs. This result highlights that not all patients are equally ready to embrace technological changes in OHC interactions. The extent to which patients can learn from ADRS varies. Previous research by (Gao et al., 2023 ) has examined the moderating role of customer ability/readiness in AI services and its impact on the value cocreation process. Our study extends this by investigating the moderation of PAR in OHCs, finding that patients with higher PAR benefit more from learning, whereas those with lower PAR struggle to process complex information. Moreover, PAR moderates the mediated relationship, which contradicts earlier studies by (Nguyen and Negash, 2024 ) and (Rebelo et al., 2024 ), who controlled for the effect of customer ability/readiness in various contexts of customer participation and citizenship behavior in online communities. However, our study shows that a patient’s desire to engage in participation behavior in OHCs significantly depends on their PAR level. For example, cancer patients with higher PAR who interact with search bots better understand the recommendations and are motivated to assist others in their recovery. Thus, our findings offer novel insights for future research on OHCs. Finally, the robustness test of the model via the MLP confirms the relevance of the integrated components, where sensitivity analysis reveals that the PAR and ADRS are the most significant predictors of VCB. Therefore, the importance of these variables should not be overlooked when aiming to achieve improved VCB outcomes. 7.2 Theoretical and Practical Implications Recent research in service studies has increasingly emphasized the importance of value cocreation, particularly with respect to the involvement of multiple actors in service environments (Wang et al., 2024 ). This study offers several significant contributions to the theoretical landscape. First, it addresses the call to explore value cocreation at the micro level(Laukka et al., 2025 ; Peng et al., 2022 ), advancing the literature by developing a sustainable environment within online health communities (OHCs). By integrating social assistance theories and value cocreation, this study proposes a moderated mediation model that lays the foundation for empirical investigations, providing substantial insights into optimizing patients' value cocreation behavior (VCB) within OHCs. Second, this research extends the application of social assistance theory to the context of human–AI interactions, considering the growing use of AI-driven tools such as AI-driven recommendation systems (ADRSs) in various sectors, including healthcare and services(Lopez-Barreiro et al., 2024 ; Varidel et al., 2025 ). The integration of AI-driven systems such as ADRS in OHCs remains an underexplored area. Our findings shed light on the different types of social assistance that patients receive from ADRS and the perceived benefits and challenges posed by these systems, which significantly influence patients' VCB. Unlike earlier studies(Liu et al., 2022 , 2020 ; Tian and Wu, 2022 ; Wang et al., 2023 ), which focused primarily on human-based social assistance, our study is the first to empirically examine how ARDS-generated support influences patient participation and contribution in OHCs, thus opening new avenues for research on how AI tools can enhance VCB in digital health contexts. Third, our study contributes to the growing literature by identifying the pivotal role of patient knowledge acquisition (PKA) in value cocreation. Peer learning has long been recognized as a key factor in value cocreation (Liu et al., 2020 ), and our study contrasts with earlier research that emphasized belongingness and ethical perceptions in facilitating value cocreation within OHCs. However, the role of PKA, particularly its empirical validation in OHCs, remains underexplored. Our findings show that PKA, facilitated through social assistance from ADRS, significantly enhances patients' VCB. Few studies have explored how PKA, as a form of SAP from AI-driven systems, mediates VCB. We show that PKA empowers patients to share information and assist others, thereby promoting value cocreation within these communities. With these results, we address the ongoing call for research into how AI search bots and recommenders can stimulate user participation in value cocreation within digital environments (Gao et al., 2023 ; Latif et al. , 2025). Fourth, this study advances our understanding of the moderating role of patient ability/readiness (PAR) in influencing PKA and VCB when interacting with the ADRS. Our findings indicate that PAR significantly enhances PKA and engagement in OHCs when patients interact with ADRSs, contributing valuable insights to the OHC literature. The proposed model suggests that while the SAP from the ADRS positively influences PKA, the strength of this relationship is contingent upon varying levels of PAR. Understanding these levels can guide the development of user-centered interventions within OHCs, fostering effective learning and collaboration for value cocreation. Finally, by combining the PLS‒SEM and MLP methods, we provide new insights that highlight the relevance of independent constructs contributing to VCB. Our results show that the MLP is a more effective predictive model, as demonstrated by the low RMSE values from both the training and testing datasets (Leong et al., 2025 ). 7.2.1. Practical Implications This study offers several practical insights for healthcare managers within online health communities (OHCs). First, the findings emphasize that ADRSs have significant potential for improving value cocreation behavior (VCB). Therefore, OHCs should go beyond using search bots solely for technical support and incorporate AI-powered ADRS that provide both emotional and informational support to community members. For example, informational support (IS) from the ADRS helps patients make informed decisions about their health by providing relevant, timely information. Additionally, ADRS can facilitate community discussions on health-related topics, allowing patients to receive diverse feedback from other members. This study further indicates that when an ADRS offers emotional support (ES), it creates empathetic responses that validate patients' feelings, encouraging them to share personal experiences and emotions within the community. While SAP from ADRS alleviates the burden on human moderators, OHCs should strike a balance between human interaction and automated support to address members’ concerns effectively. Second, the study revealed that patient knowledge acquisition (PKA) mediates the relationship between ADRS and VCB. This suggests that OHC managers should focus on patient education when AI-driven systems are integrated. By offering personalized resources and information tailored to patients' health needs, OHCs can enhance PKA. For example, personalized guidance from the ADRS, which responds to specific health-related queries, boosts patients' confidence in their ability to participate in discussions and advocate for others within the community. Third, our findings underscore that SAP from ADRSs plays a critical role in fostering PKA and promoting value cocreation within OHCs. However, the effectiveness of ADRS may vary depending on patient ability/readiness (PAR). OHC managers should account for different levels of PAR when designing AI-based support systems. By assessing PAR levels and adopting personalized approaches, search bot responses can be tailored to patients' comprehension levels. The complexity of the language used by ADRS affects how well patients understand and engage with the information. Healthcare managers should ensure that ADRS simplify complex health-related content. Furthermore, OHCs should create safe spaces where patients with lower PAR feel comfortable seeking clarification about their health issues. Patients with higher PAR are more likely to engage with ADRS and actively participate in value cocreation activities within the community. Finally, this study illustrates how the indirect effect of the SAP from the ADRS on VCB through the PKA is influenced by the PAR. Specifically, the SAP from the ADRS significantly impacts VCB through the PKA when the PAR is high. However, this indirect effect becomes nonsignificant when PAR is low. The identification of this boundary condition provides a deeper understanding of the relationship between ADRS and patients' abilities, contributing to the social assistance literature. These findings also extend current research by highlighting the conditional indirect pathway through which SAPs from the ADRS, PKA, and PAR collectively influence VCB. 8. Conclusion and Limitations This study demonstrates that AI-driven recommendation systems (ADRS) significantly enhance value cocreation behavior (VCB) within online health communities (OHCs) by fostering patient learning and promoting trust-based engagement. The findings reveal that trust in ADRS is the most critical determinant of both learning and cocreation, with perceived risk and privacy concerns acting as barriers to participation. Furthermore, health anxiety positively moderates the relationship between trust and learning, and patient readiness (PAR) strengthens the indirect effect of trust on VCB through learning. By integrating PLS-SEM and MLP, we provide a robust analytical framework that captures both causal and nonlinear relationships, ensuring the reliability and generalizability of the model. From a theoretical perspective, this research extends the understanding of AI-enabled value cocreation by integrating insights from trust theory, learning mechanisms, and the privacy calculus framework within healthcare contexts. Practically, the findings suggest that AI systems designed with transparency, privacy assurance, and emotional sensitivity can significantly enhance patient engagement and collaborative behaviors. Healthcare organizations and ADRS developers should focus on improving user trust and readiness while ensuring that AI tools are empathetic, explainable, and privacy-conscious, empowering patients to actively participate in their digital health journeys. However, several limitations should be considered. First, the study relied on reported data from questionnaire-based surveys, and future research could benefit from incorporating objective data directly from patient interactions within OHCs to improve accuracy. Second, while this study focused on OHCs in China, the findings may not be generalizable to other regions. Future studies should examine the model in diverse geographical contexts to assess its global applicability. Additionally, this study focused primarily on younger patients, so exploring age-related differences in VCB would offer valuable insights. Finally, while PLS-SEM and MLP were effective in this study, future research could explore other machine learning techniques to increase the predictive accuracy and gain a more nuanced understanding of the factors influencing VCB within OHCs. These limitations open several avenues for future research, and addressing them would further contribute to advancing the understanding of ADRS in healthcare and its role in fostering VCB. Declarations Conflicts of Interest: The authors declare no conflict of interest. Funding: The authors received no financial support for the research and/or authorship of this article. Data Availability The data used in this study are available from the corresponding author upon reasonable request. References Aggarwal, C.C. (2015), “Outlier Analysis”, Data Mining , Springer International Publishing, pp. 237–263, doi: 10.1007/978-3-319-14142-8_8. Ahuja, H. and Shree, D. (2025), “Patient-to-patient value co-creation in online health communities: a systematic literature review and research agenda”, International Journal of Pharmaceutical and Healthcare Marketing , Emerald Publishing, Vol. 19 No. 1, pp. 107–138, doi: 10.1108/IJPHM-05-2024-0046. Ain, F.U., Siddiqui, A., Mustafa, A., Memon, F. and Kausar, F. (2025), “AI-DRIVEN DECISION SUPPORT SYSTEMS IN PUBLIC HEALTH ADMINISTRATION”, Journal of Medical & Health Sciences Review , Insightful Education Research Institute, Vol. 2 No. 3, doi: 10.62019/WAPQJ864. Bagheri, M., Bagheritabar, M., Alizadeh, S., Parizi, M. (Sam) S., Matoufinia, P. and Luo, Y. (2025), “Machine-Learning-Powered Information Systems: A Systematic Literature Review for Developing Multi-Objective Healthcare Management”, Applied Sciences (Switzerland) , Multidisciplinary Digital Publishing Institute (MDPI), 1 January, doi: 10.3390/app15010296. Bahit, M.C., Korjian, S., Chi, G., Daaboul, Y., Castro Molina, A., Jiang, G., Dangas, C., et al. (2025), “Understanding the Burden of Myocardial Infarction and Patient Preferences for Treatment: A Real-World Study Assessing Patients’ Perspectives via an Online Survey”, Journal of Patient-Centered Research and Reviews , Aurora Health Care, Inc., Vol. 12 No. 4, pp. 234–244, doi: 10.17294/2330-0698.2146. Chen, L., Baird, A. and Straub, D. (2019), “Fostering participant health knowledge and attitudes: an econometric study of a chronic disease-focused online health community”, Journal of Management Information Systems , Taylor & Francis, Vol. 36 No. 1, pp. 194–229. Dralus, G., Mazur, D., Kusznier, J. and Dralus, J. (2023), “Application of artificial intelligence algorithms in multilayer perceptron and Elman networks to predict photovoltaic power plant generation”, Energies , MDPI, Vol. 16 No. 18, p. 6697. Febri Ramanda, M. Ari Prayogo, Bagus Dwi Saputra and Muhammad Labib Jundillah. (2025), “Exploring User Experience in Adopting AI-Based Information Systems in Healthcare Environments”, Jurnal Ilmiah Sistem Informasi Dan Ilmu Komputer , Politeknik Pratama Purwokerto, Vol. 5 No. 2, pp. 517–529, doi: 10.55606/JUISIK.V5I2.1437. Gao, B., Wang, J., Ding, X. and Guo, Y. (2024), “The Pitfalls of Review Solicitation: Evidence from a Natural Experiment on TripAdvisor”, Https://Doi.Org/10.1287/Mnsc.2023.01006 , INFORMS, Vol. 71 No. 2, pp. 1671–1691, doi: 10.1287/MNSC.2023.01006. Gao, L., Li, G., Tsai, F., Gao, C., Zhu, M. and Qu, X. (2023), “The impact of artificial intelligence stimuli on customer engagement and value co-creation: the moderating role of customer ability readiness”, Journal of Research in Interactive Marketing , Emerald Publishing, Vol. 17 No. 2, pp. 317–333, doi: 10.1108/JRIM-10-2021-0260. Glikson, E. and Woolley, A.W. (2020), “Human Trust in Artificial Intelligence: Review of Empirical Research”, Https://Doi.Org/10.5465/Annals.2018.0057 , Academy of Management AnnalsBriarcliff Manor, NY, Vol. 14 No. 2, pp. 627–660, doi: 10.5465/ANNALS.2018.0057. Grassini, E., Buzzi, M., Leporini, B. and Vozna, A. (2025), “A systematic review of chatbots in inclusive healthcare: insights from the last 5 years”, Universal Access in the Information Society , Springer, Vol. 24 No. 1, pp. 195–203. Guan, R., Rakovic, M., Chen, G. and Gaševic, D. (2025), “How educational chatbots support self-regulated learning? A systematic review of the literature”, Education and Information Technologies , Springer, Vol. 30 No. 4, pp. 4493–4518. Hair, J. and Alamer, A. (2022), “Partial Least Squares Structural Equation Modeling (PLS-SEM) in second language and education research: Guidelines using an applied example”, Research Methods in Applied Linguistics , Elsevier, Vol. 1 No. 3, p. 100027. Hair, J.F., Risher, J.J., Sarstedt, M. and Ringle, C.M. (2019), “When to use and how to report the results of PLS-SEM”, European Business Review , Emerald Publishing Limited, Vol. 31 No. 1, pp. 2–24. Hair, J.F., Singh, S., Kaur, R. and Dana, L.P. (2022), “A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM)”, SAGE Publication , SAGE Publication, Inc., pp. 1–352. Harkness, J.A., Braun, M., Edwards, B., Johnson, T.P., Lyberg, L.E., Mohler, P.P., Pennell, B.-E., et al. (2010), Survey Methods in Multinational, Multiregional, and Multicultural Contexts , John Wiley & Sons. Henseler, J., Ringle, C.M. and Sarstedt, M. (2015), “A new criterion for assessing discriminant validity in variance-based structural equation modeling”, Journal of the Academy of Marketing Science , Springer, Vol. 43 No. 1, pp. 115–135. Jannach, D. and Adomavicius, G. (2016), “Recommendations with a purpose”, RecSys 2016 - Proceedings of the 10th ACM Conference on Recommender Systems , Association for Computing Machinery, Inc, pp. 7–10, doi: 10.1145/2959100.2959186;JOURNAL:JOURNAL:ACMCONFERENCES;PAGEGROUP:STRING:PUBLICATION. Kadir, I., Abdallaoui, A. and Elazhari, K. (2025), “Architectural optimization of a multilayer perceptron (MLP) neural network enhanced by the Levenberg - - Marquardt algorithm for predicting relative humidity: application to Tangier, Morocco”, Water SA , Water Research Commission (WRC), Vol. 51 No. 3, pp. 279–287. Ketchen, D.J. (2013), “A Primer on Partial Least Squares Structural Equation Modeling”, Long Range Planning , Elsevier BV, Vol. 46 No. 1–2, pp. 184–185, doi: 10.1016/j.lrp.2013.01.002. Kock, N. (2015), “Common method bias in PLS-SEM: A full collinearity assessment approach”, International Journal of E-Collaboration (Ijec) , IGI Global, Vol. 11 No. 4, pp. 1–10. Kock, N. (2020), “Harman’s single factor test in PLS-SEM: Checking for common method bias”, Data Analysis Perspectives Journal , Vol. 2 No. 2, pp. 1–6. Al Kuwaiti, A., Nazer, K., Al-Reedy, A., Al-Shehri, S., Al-Muhanna, A., Subbarayalu, A.V., Al Muhanna, D., et al. (2023), “A review of the role of artificial intelligence in healthcare”, Journal of Personalized Medicine , MDPI, Vol. 13 No. 6, p. 951. Latan, H., Hair Jr, J.F., Noonan, R. and Sabol, M. (2023), “Introduction to the partial least squares path modeling: Basic concepts and recent methodological enhancements”, Partial Least Squares Path Modeling: Basic Concepts, Methodological Issues and Applications , Springer, pp. 3–21. Latif, M.S. and Wang, J.-J. (2025), “The moderating role of face on value co-creation behavior and co-creation attitude in online health communities”, Aslib Journal of Information Management , Emerald Publishing Limited, Vol. 77 No. 3, pp. 464–487. Latif, M.S., Wang, J.J., Shahzad, M. and Mursil, M. (2024), “The impact of social support chatbots on patients’ value co-creation behavior in online health communities: a moderated mediation model”, Internet Research , Emerald Publishing Limited, Vol. 35 No. 3, pp. 1126–1153, doi: 10.1108/INTR-08-2023-0666. Laukka, E., Tuunanen, T., Jansson, M., Vanhanen, M., Hirvonen, N., Palukka, J., Vesinurm, M., et al. (2025), “Value Cocreation and Codestruction in Digital Health Services: Protocol for a Systematic Review”, JMIR Research Protocols , JMIR Publications Inc., Toronto, Canada, Vol. 14 No. 1, p. e63015. Leong, L.-Y., Hew, T.-S., Ooi, K.-B., Tan, G.W.-H. and Koohang, A. (2025), “An SEM-ANN approach-guidelines in information systems research”, Journal of Computer Information Systems , Taylor & Francis, Vol. 65 No. 6, pp. 706–737. Leskovec, Jurij., Rajaraman, Anand. and Ullman, J.D.. (2020), “Mining of massive datasets”, Cambridge University Press, p. 553. Li, L., Pongsakornrungsilp, S., Pongsakornrungsilp, P. and Jiang, Y. (2024), “The Influence of Social Support on Customer Engagement in Live Streaming Commerce: The Mediating Role of Perceived Value.”, TEM Journal , Vol. 13 No. 4. Li, M. and Tuunanen, T. (2022), “Information Technology–Supported value Co-Creation and Co-Destruction via social interaction and resource integration in service systems”, The Journal of Strategic Information Systems , Elsevier, Vol. 31 No. 2, p. 101719. Liu, M., Zou, X., Chen, J. and Ma, S. (2022), “Comparative analysis of social support in online health communities using a word co-occurrence network analysis approach”, Entropy , MDPI, Vol. 24 No. 2, p. 174. Liu, S., Xiao, W., Fang, C., Zhang, X. and Lin, J. (2020), “Social support, belongingness, and value co-creation behaviors in online health communities”, Telematics and Informatics , Elsevier, Vol. 50, p. 101398. Lopez-Barreiro, J., Garcia-Soidan, J.L., Alvarez-Sabucedo, L. and Santos-Gago, J.M. (2024), “Artificial intelligence-powered recommender systems for promoting healthy habits and active aging: A systematic review”, Applied Sciences , MDPI, Vol. 14 No. 22, p. 10220. Miner, A.S., Laranjo, L. and Kocaballi, A.B. (2020), “Chatbots in the fight against the COVID-19 pandemic”, NPJ Digital Medicine , Nature Publishing Group UK London, Vol. 3 No. 1, p. 65. Miyazaki, A.D. and Haderlie, T.C. (2025a), “Emotional AI as a facilitator of co-created service value in a health-care ecosystem: a collective framework”, Journal of Services Marketing , Emerald Publishing Limited, doi: 10.1108/JSM-01-2025-0041. Naga, J., Quidato, M., Pascual, Ma.C.G.P. and Ebardo, R. (2025), “Engagement Dynamics in Online Mental Health Communities: The Role of Trust and Social Support in Young Adults’ Attitudes and Norms”, SAGE Open , SAGE Publishing, Vol. 15 No. 3, doi: 10.1177/21582440251376147. Nguyen, T.T.V. and Negash, Y.T. (2024), “The Impact of Value Co-Creation Readiness on Customer Experience and Revisit Intention in the Hospitality Industry.”, International Journal of Information & Management Sciences , Vol. 35 No. 3. Nohutlu, Z.D., Englis, B.G., Groen, A.J. and Constantinides, E. (2023), “Innovating with the customer: Co-creation motives in online communities”, International Journal of Electronic Commerce , Taylor & Francis, Vol. 27 No. 4, pp. 523–557. Olawade, D.B., Wada, O.Z., Odetayo, A., David-Olawade, A.C., Asaolu, F. and Eberhardt, J. (2024), “Enhancing mental health with Artificial Intelligence: Current trends and future prospects”, Journal of Medicine, Surgery, and Public Health , Elsevier, Vol. 3, p. 100099, doi: 10.1016/J.GLMEDI.2024.100099. Osei-Frimpong, K., Wilson, A. and Lemke, F. (2018), “Patient co-creation activities in healthcare service delivery at the micro level: The influence of online access to healthcare information”, Technological Forecasting and Social Change , Elsevier, Vol. 126, pp. 14–27. “(PDF) Collaborative Filtering Recommender Systems”. (n.d.)., available at: https://www.researchgate.net/publication/200121027_Collaborative_Filtering_Recommender_Systems (accessed 9 December 2025). Peng, Y., Wu, T., Chen, Z. and Deng, Z. (2022), “Value cocreation in health care: systematic review”, Journal of Medical Internet Research , JMIR Publications Toronto, Canada, Vol. 24 No. 3, p. e33061. Podsakoff, P.M., Podsakoff, N.P., Williams, L.J., Huang, C. and Yang, J. (2024), “Common method bias: It’s bad, it’s complex, it’s widespread, and it’s not easy to fix”, Annual Review of Organizational Psychology and Organizational Behavior , Annual Reviews, Vol. 11 No. 1, pp. 17–61. Qiu, X., Sun, T., Xu, Y., Shao, Y., Dai, N. and Huang, X. (2021), “Pre-trained Models for Natural Language Processing: A Survey”, Science China Technological Sciences , Springer Verlag, Vol. 63 No. 10, pp. 1872–1897, doi: 10.1007/s11431-020-1647-3. Ramadan, O.M.E., Elsharkawy, N.B., Hafiz, A.H., Katooa, N.E., Abunar, A., Attallah, D.M.A., Dewan, M., et al. (2025), “Neonatal nurses’ e-health literacy and technology-mediated clinical practice: a cross-sectional analysis of digital health competencies and practice patterns”, BMC Nursing , BioMed Central, Vol. 24 No. 1, doi: 10.1186/S12912-025-03839-7. Rebelo, A., Varela-Neira, C. and Ruzo-Sanmartín, E. (2024), “Boosting Customers’ Co-Creation in Open-Source Software Environments: The Role of Innovativeness and a Sense of Community”, Journal of Theoretical and Applied Electronic Commerce Research , MDPI, Vol. 19 No. 3, pp. 2476–2496. Ricci, F., Rokach, L. and Shapira, B. (2022), “Recommender Systems Handbook: Third Edition”, Recommender Systems Handbook: Third Edition , Springer US, pp. 1–1060, doi: 10.1007/978-1-0716-2197-4/COVER. Ringle, C.M., Wende, S. and Becker, J. (2024), “SmartPLS 4. Bönningstedt: SmartPLS. http s”. Roppelt, J.S., Jenkins, A., Kanbach, D.K., Kraus, S. and Jones, P. (2025), “Effective adoption of artificial intelligence in healthcare: A multiple case study”, Journal of Decision Systems , Taylor & Francis, Vol. 34 No. 1, p. 2458883. Sanaji, S., Handriana, T. and Usman, I. (2022), “Antecedents and consequences of value co-creation in online shopping”, BISMA (Bisnis Dan Manajemen) , Universitas Negeri Surabaya, Vol. 15 No. 1, pp. 20–39, doi: 10.26740/BISMA.V15N1.P20-39. Schroer, H.W. and Just, C.L. (2023), “Feature engineering and supervised machine learning to forecast biogas production during municipal anaerobic co-digestion”, ACS Es&t Engineering , ACS Publications, Vol. 4 No. 3, pp. 660–672. Sharma, A., Lin, I.W., Miner, A.S., Atkins, D.C. and Althoff, T. (2023), “Human–AI collaboration enables more empathic conversations in text-based peer-to-peer mental health support”, Nature Machine Intelligence , Nature Publishing Group UK London, Vol. 5 No. 1, pp. 46–57. Swan, E.L., Peltier, J.W. and Dahl, A.J. (2024), “Artificial intelligence in healthcare: the value co-creation process and influence of other digital health transformations”, Journal of Research in Interactive Marketing , Emerald Publishing Limited, Vol. 18 No. 1, pp. 109–126. Szeto, M.D., Hook Sobotka, M., Woolhiser, E., Parmar, P., Wu, J., Alhanshali, L. and Dellavalle, R.P. (2024), “PatientsLikeMe and online patient support communities in dermatology”, JMIR Dermatology , JMIR Publications Toronto, Canada, Vol. 7, p. e50453. Tan, Q., Tan, J. and Gao, X. (2024), “How does the online innovation community climate affect the user’s value co-creation behavior: The mediating role of motivation”, PLOS ONE , Public Library of Science, Vol. 19 No. 4, p. e0301299, doi: 10.1371/JOURNAL.PONE.0301299. Tao, Guannan; Yan, Sanjiu, ; and Zhou, Y. (2025), “Analysis of the Driving Factors for Fan economy Growth on Social Media Platforms Based on the Push-Pull Theory: Taking the Largest Platform in China as an Example Standard-Nutzungsbedingungen”. The Annenberg Public Policy Center of the University of Pennsylvania. (2025), “Many in U.S. Consider AI-Generated Health Information Useful and Reliable | The Annenberg Public Policy Center of the University of Pennsylvania”, July, available at: https://www.annenbergpublicpolicycenter.org/many-in-u-s-consider-ai-generated-health-information-useful-and-reliable/ (accessed 2 December 2025). Tian, X.-F. and Wu, R.-Z. (2022), “Determining factors affecting the users’ participation of online health communities: an integrated framework of social capital and social support”, Frontiers in Psychology , Frontiers Media SA, Vol. 13, p. 823523. Trigka, M. and Dritsas, E. (2025), “A comprehensive survey of deep learning approaches in image processing”, Sensors , MDPI, Vol. 25 No. 2, p. 531. Vanderhout, S., Smith, M., Lee-Foon, N., Doherty-Kirby, A., Fleming, R., Grant, D., Grenier, A., et al. (2025), “Co-creating an everyday language illustration of learning health systems alongside patient, caregiver, and community partners”, Learning Health Systems , Wiley, doi: 10.1002/LRH2.70033. Varidel, M., An, V., Hickie, I.B., Cripps, S., Marchant, R., Scott, J., Crouse, J.J., et al. (2025), “Causal AI Recommendation System for Digital Mental Health: Bayesian Decision-Theoretic Analysis”, Journal of Medical Internet Research , JMIR Publications Toronto, Canada, Vol. 27, p. e71305. Vickery, S.K., Castel, M.J., Narayanan, S. and Nicolae, M.L. (2025), “A Model of Hospital Patient Engagement for Value Co-Creation: Does It Affect Performance? A U.S. Hospital Industry Study”, Journal of Operations Management , Wiley, doi: 10.1002/JOOM.70019. Vinh, V.M. and Hung, N.G. (2025), “AI-Driven Success: Exploring the Role of Technology Readiness and Artificial Intelligence Integration on Digital Business Performance”, Journal of Economics, Finance and Management Studies , Everant Journals, Vol. 08 No. 09, doi: 10.47191/JEFMS/V8-I9-44. Wang, C., Zhao, X. and Hong, J. (2024), “A Meta-Analysis of the Effects of Interaction on Value Co-Creation in Online Collaborative Innovation Communities Based on the Service Ecosystem Framework”, Behavioral Sciences , MDPI, Vol. 14 No. 12, p. 1177. Wang, J., Yao, T. and Wang, Y. (2023), “Patient engagement as contributors in online health communities: the mediation of peer involvement and moderation of community status”, Behavioral Sciences , MDPI, Vol. 13 No. 2, p. 152. Wei, J. (2025a), “Analysis of the Behavioral Performance and Social Support of Patients in Online Health Communities From User Profile Perspectives: Comparative Study”, Journal of Medical Internet Research , JMIR Publications Toronto, Canada, Vol. 27, p. e68074. Xiao Fei, Y., Loh, X.M., Seduram, L.A. /P and Tan, G.W.H. (2025), “How is brand love shaped through virtual event value co-creation and sponsorship? A multi-method approach”, Journal of Research in Interactive Marketing , Emerald Publishing, doi: 10.1108/JRIM-11-2024-0516/1255978. Xiao, L. and Han, X. (2025), “Enhancing Patient Value Co-creation via AI-enabled Cases Disclosure: The Role of Uncertainty Reduction, Patient Empowerment and Risk Aversion”, Social Science & Medicine , Elsevier, p. 118822. Yang, B., Lu, W., Xuan, Y., Hao, C. and Huang, X. (2025), “The influences of social support expressed from doctors and disclosed from peers on patient decision-making: an analysis from the online health community”, Dental Science Reports , Nature Portfolio, Vol. 15 No. 1, doi: 10.1038/S41598-024-85023-6. Yang, J., Ren, J., Zhao, M., Chen, H. and Wang, H. (2023), “The impact of value co-creation on consumer citizenship behavior: Based on consumer perspective”, Frontiers in Psychology , Frontiers Media S.A., Vol. 13, p. 1110011, doi: 10.3389/FPSYG.2022.1110011/BIBTEX. Yi, Y. and Gong, T. (2013), “Customer value co-creation behavior: Scale development and validation”, Journal of Business Research , Elsevier, Vol. 66 No. 9, pp. 1279–1284, doi: 10.1016/J.JBUSRES.2012.02.026. Yin, J., Zhao, Y., Ma, Z., Chen, S. and Li, M. (2025), “Customer citizenship behavior and customer perceived value in China: the mediating role of value co-creation experience”, Humanities and Social Sciences Communications , Palgrave. Zhang, Z. and Lu, M. (2025), “When engagement meets departure: how social capital, social engagement and social support impact churn within online health communities”, Aslib Journal of Information Management , Emerald Publishing Limited, doi: 10.1108/AJIM-04-2024-0350. Zhao, H. (2025), “The impact of social support on individual health knowledge creation in online patient communities: the mediating role of attitude”, Journal of Knowledge Management , Emerald Publishing Limited, Vol. 29 No. 3, pp. 937–967. Zhao, L. and Liu, T. (2025), “Exploring generational and educational disparities in online health information seeking in China: The moderating role of internet use”, Digital Health , SAGE Publications Inc., Vol. 11, doi: 10.1177/20552076251393378. Zhou, P., Zhao, Y., Xiao, S. and Zhao, K. (2022), “The impact of online health community engagement on lifestyle changes: A serially mediated model”, Frontiers in Public Health , Frontiers Media SA, Vol. 10, p. 987331. Zou, J. and Shao, Y. (2022), “A Study on Factors Affecting the Value Co-Creation Behavior of Customers in Sharing Economy: Take Airbnb Malaysia as an Example”, Sustainability 2022, Vol. 14, Page 12678 , Multidisciplinary Digital Publishing Institute, Vol. 14 No. 19, p. 12678, doi: 10.3390/SU141912678. Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8413621","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":563469016,"identity":"8fcdbe6f-5cbb-42bd-a513-6ca6d9e99305","order_by":0,"name":"Sadia Shabbir","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA80lEQVRIiWNgGAWjYBACfv7mgw8SDGzkkMQS8GuRnHEs2eBDRZoxAxuxWgwO5JhJzjhzOLGBaC0MBw4YG/O2HU6fP7/HgLmg4jADP3uOAcPPHbh1MDY3JD7mbUvP3XCMx4AZaB2DZM8bA8beM7i1MDMcOAy0xTp3AxtQC9A6BoMbOQbMjG24tbAxJLZJ87Yxp8u3gbT8O8xgT0gLD0MyG9D7zgkMIIfxNgBtkSCgRULiGDMokA03HEsrOMxzLJ1H4syzgoO9eLTYn+//CIpKefnmwxsf89RYy/G3J2988BOPFhRwAOxSKGMUjIJRMApGAQUAAIfmT3Zxf/1AAAAAAElFTkSuQmCC","orcid":"","institution":"Northeastern University","correspondingAuthor":true,"prefix":"","firstName":"Sadia","middleName":"","lastName":"Shabbir","suffix":""},{"id":563469017,"identity":"fae11911-e65e-480f-adf8-4bfc06701d24","order_by":1,"name":"Qiu Rouzhen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3klEQVRIiWNgGAWjYJCCAwwVDAxs7AguMVrOALUwk6KFgbENSBCtxbz/8MbDvPO2yfMxM7A9+NjGIMd3I4HxcwEeLTIHjhUc5t1227CNmYHdcGYbg7HkjQRm6Rl4tEgw9hiAtDACtbBJ87YxJG64kcDGzINPCzMPUMuc2/ZgLX/bGOoJa2EDaWm4nQjWAgyHBAOCWnjYCg7OOXY7uY2ZsU2y55yE4cwzD5ul8WrhP7z5w5ua27bz25uPSfwos5HnO5588DM+LUBgAKUZG0BGwBhEaRkFo2AUjIJRgAMAAME3Q+jXEEjTAAAAAElFTkSuQmCC","orcid":"","institution":"Northeastern University","correspondingAuthor":true,"prefix":"","firstName":"Qiu","middleName":"","lastName":"Rouzhen","suffix":""},{"id":563469018,"identity":"baafe0f1-d424-467b-aec3-ddc8d5cc1b27","order_by":2,"name":"Naseeb Khan","email":"","orcid":"","institution":"Northeastern University","correspondingAuthor":false,"prefix":"","firstName":"Naseeb","middleName":"","lastName":"Khan","suffix":""},{"id":563469019,"identity":"597fafb2-f2b3-401b-a6ec-96cb4fd9c211","order_by":3,"name":"Muhammad Usman Arshad","email":"","orcid":"","institution":"Dalian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Muhammad","middleName":"Usman","lastName":"Arshad","suffix":""}],"badges":[],"createdAt":"2025-12-20 17:28:13","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-8413621/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8413621/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99191273,"identity":"8a05d3df-3c48-4829-816e-31b217695de5","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":692420,"visible":true,"origin":"","legend":"","description":"","filename":"reviewedManuscriptV1SF1V2.docx","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/1da456b28e42dadd78b2b16f.docx"},{"id":99191272,"identity":"f355525b-20bf-46de-a22b-05aa6a601685","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":342,"visible":true,"origin":"","legend":"","description":"","filename":"rs8413621.json","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/def7b39c2eab7311410080bb.json"},{"id":99191277,"identity":"774d6890-6ded-4a92-97ee-727bbf0451a6","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":242658,"visible":true,"origin":"","legend":"","description":"","filename":"rs84136210enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/54f65f131257e6560bceaf28.xml"},{"id":99191281,"identity":"cba48aed-37a2-4af9-8df4-c7c7df9867f2","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":49241,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/0c1f96ac4f31b964a90aede6.png"},{"id":99191285,"identity":"e28af350-1f0f-4e3c-9d7d-598a185850fd","added_by":"auto","created_at":"2025-12-30 00:54:43","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":78988,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/a2aa32c202cfa851e93d9339.png"},{"id":99191284,"identity":"7154b76d-0933-457d-b7b2-2cce4545651a","added_by":"auto","created_at":"2025-12-30 00:54:43","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":114222,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/25a557ecc611588d2835a6ed.png"},{"id":99191287,"identity":"765c248b-d0ef-4eeb-b641-5c45afefb95d","added_by":"auto","created_at":"2025-12-30 00:54:43","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25772,"visible":true,"origin":"","legend":"","description":"","filename":"groupimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/1026faa6056d956e222d2871.jpeg"},{"id":99191275,"identity":"40812634-12cf-44bd-b8ed-6aa1f308f357","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17585,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/d27be960b434b747d0fb40f8.png"},{"id":99191282,"identity":"516580bd-89b3-4b1c-a665-e1782821f61d","added_by":"auto","created_at":"2025-12-30 00:54:43","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":18127,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/97ccca76a3fd88efbaebdfd6.png"},{"id":99191283,"identity":"aa6f1d40-8c3a-4538-8590-dd33bdc57946","added_by":"auto","created_at":"2025-12-30 00:54:43","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":30262,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/11cd601189bec3206b20e547.png"},{"id":99191280,"identity":"0da8b741-2903-4913-9ecd-f801125de66d","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6054,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinegroupimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/0a617efaa794282a77640236.png"},{"id":99191286,"identity":"168c31c2-3446-40db-a7e3-568ed396af4f","added_by":"auto","created_at":"2025-12-30 00:54:43","extension":"xml","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":238220,"visible":true,"origin":"","legend":"","description":"","filename":"rs84136210structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/8f76059135932784cabb021a.xml"},{"id":99191279,"identity":"35c4697f-0c42-466e-8929-b11c888392a6","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"html","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":255510,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/ab721594275f153d1440f5e1.html"},{"id":99191271,"identity":"2995549c-a9d4-4b24-b8c2-ca3f717e4dc3","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":29016,"visible":true,"origin":"","legend":"\u003cp\u003eResearch Framework/Model\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/0a51299e35cf441e4053d3da.png"},{"id":99191274,"identity":"71314810-079b-444a-bfe4-7bd7f497f69d","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":60025,"visible":true,"origin":"","legend":"\u003cp\u003eBrief overview of the methodology flowchart\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/642442b998f34c53d003b2b3.png"},{"id":99317260,"identity":"038e528b-2de7-45e8-9a5a-5df9a1b8e6f0","added_by":"auto","created_at":"2025-12-31 16:29:51","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":145657,"visible":true,"origin":"","legend":"\u003cp\u003eVisualization of data characteristics\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/055361200c7e76ea1342216f.png"},{"id":99191278,"identity":"7baeb5d2-4659-4f44-8323-f22a38a84337","added_by":"auto","created_at":"2025-12-30 00:54:42","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":171042,"visible":true,"origin":"","legend":"\u003cp\u003eInteraction effect of social assistance perceived (SAP) from AI-driven recommendation systems and patient ability/readiness on patient knowledge acquisition.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/e01852f302f666435690ff91.png"},{"id":99323653,"identity":"b16be339-2838-4dba-b703-53515d3d73e3","added_by":"auto","created_at":"2025-12-31 16:45:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2306494,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8413621/v1/31e73664-5b32-4119-a97d-cb947777833c.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eThe Impact of AI-Driven Recommendation Systems on Patients’ Value Co-Creation Behavior in Online Health Communities: A Moderated Mediation Model\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe rapid evolution of information technology has significantly transformed healthcare delivery, particularly through the growing use of online health communities (OHCs), which have become pivotal in modern healthcare systems. In the modern Era, with fast pacing life, where people require quick answers to their queries, AI-based recommenders play a vital role in fetching the most relevant search or comments while searching online communities. In the United States, recent survey data show that approximately 79% of adults visit online to look up answers about health symptoms or conditions, and approximately 71% use search engines such as Google or Bing at least occasionally to find health information(The Annenberg Public Policy Center of the University of Pennsylvania, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). According to (Zhao and Liu, \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), 78.6% of people were using the internet in China by December 2024, and approximately 418\u0026nbsp;million of them had searched for health information and medical services online. The study also highlighted that people frequently use the internet and smartphones to look up health-related information across South Asian countries, including Vietnam, Indonesia, India, the Philippines, Singapore, and Japan. OHCs, such as PatientsLikeMe.com (Szeto et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and Carenity.com(Bahit et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), have gained popularity, providing patients with a platform to exchange health-related experiences, ask questions, and receive feedback from peers. These communities offer valuable opportunities for patients to collaborate, share health-specific information, and access crucial social assistance (e.g., emotional and informational aid), thereby enhancing patient self-efficacy and bridging gaps in traditional care delivery (Wei, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2025a\u003c/span\u003e; Zhou et al., \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, even with their increasing usage, the success and long-term viability of OHCs are often hindered by a lack of patient engagement and contribution. The critical question remains: how can OHCs sustain patient participation over time?\u003c/p\u003e \u003cp\u003eResearch has shown that value cocreation behavior (VCB) plays a crucial role in maintaining OHCs (Latif and Wang, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Traditionally, patients have often been regarded as passive recipients of healthcare services, a model rooted in a one-way, provider-dominated communication flow that severely limits the patient\u0026rsquo;s capacity to contribute to the overall value creation process (Peng et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, the emergence of OHCs has shifted this dynamic, allowing patients to actively participate in knowledge sharing and providing feedback on their health experiences. These interactions, particularly patient-to-patient and patient-to-physician interactions, have been identified as key drivers of VCB (Latif and Wang, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Osei-Frimpong et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Recent advancements in artificial intelligence (AI) have also spurred the adoption of AI-driven recommendation systems (ADRS), including relevant comment fetchers, point search results and social assistance search bots (SSBs), in healthcare settings (Bagheri et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Gao et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Grassini et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). These AI tools are increasingly used to provide medical guidance, answer patient queries, and support health management (Al Kuwaiti et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and have demonstrated their potential during crises such as the COVID-19 pandemic, where recommender systems have played a significant role in reducing the workload of healthcare professionals (Miner et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn addition to providing technical assistance, AI-based recommenders also provide emotional and informational support, often fostering empathetically and collaborative practices (Sharma et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This shift toward AI-driven support systems in OHCs highlights the importance of understanding how these tools influence patient behavior in the cocreation process. While prior research has explored the role of human-based social assistance in encouraging VCB (Latif and Wang, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), there remains a gap in understanding how AI-powered tools such as ADRS can increase patient participation and engagement. Despite the growing use of recommenders and search bots, the mechanisms through which these AI tools influence patients\u0026rsquo; collaborative behavior in OHCs are still not well understood.\u003c/p\u003e \u003cp\u003eThis study aims to fill this gap by examining the role of ADRS in driving VCB within OHCs. Drawing on service-dominant logic (SDL) and value cocreation theory (VCT), this study explores how AI tools such as recommenders and search bots act as resource integrators that encourage VCB in these communities. Moreover, it examines the mediating role of patient knowledge acquisition (PKA) and investigates how patient ability/readiness (PAR) moderates the relationship between social assistance perceived (SAP) from ADRS and VCB. The findings of this study will contribute to the growing body of research on AI in healthcare by providing insights into the role of ADRS in enhancing patient engagement, improving health outcomes, and fostering sustainable OHCs. This study also responds to recent calls for empirical research on how artificial intelligence enables value cocreation in healthcare and digital health ecosystems (Roppelt et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Swan et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Xiao and Han, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and offers actionable implications for healthcare managers seeking to strengthen patient participation and value cocreation behaviors in online health communities (Liu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe study makes three main contributions. Theoretically, it enriches the understanding of AI-enabled value cocreation by integrating trust, learning, and emotional readiness constructs within a single conceptual model. Methodologically, it introduces a hybrid PLS-SEM\u0026ndash;MLP framework that merges the strengths of structural modelling and deep learning-based prediction. Practically, it offers insights for AI developers and healthcare institutions in designing systems that foster trust, empathy, and privacy assurance, thereby enhancing patient participation and digital health literacy. Through this integrated perspective, this study aims to advance both the scholarly understanding and practical implementation of AI-driven engagement in healthcare communities.\u003c/p\u003e \u003cp\u003eThe remainder of this paper is structured as follows. Section 2 develops the theoretical framework and formulates the research hypotheses that form the foundation of the study. Section \u003cspan refid=\"Sec6\" class=\"InternalRef\"\u003e3\u003c/span\u003e describes the research framework. Section \u003cspan refid=\"Sec10\" class=\"InternalRef\"\u003e4\u003c/span\u003e describes the methodology, including the data collection procedures, sampling details, and measurement instruments. Section \u003cspan refid=\"Sec14\" class=\"InternalRef\"\u003e5\u003c/span\u003e presents the data analysis methods, validation tests via PLS‒SEM, and empirical findings, including hypothesis testing, mediation, and moderation analyses. Section \u003cspan refid=\"Sec19\" class=\"InternalRef\"\u003e6\u003c/span\u003e details the robustness checks conducted through multilayer perceptron (MLP) modelling. Section \u003cspan refid=\"Sec20\" class=\"InternalRef\"\u003e7\u003c/span\u003e discusses the results in light of the literature, outlining the theoretical and practical implications. Section \u003cspan refid=\"Sec24\" class=\"InternalRef\"\u003e8\u003c/span\u003e concludes the study by summarizing key findings, acknowledging limitations, and suggesting directions for future research. Collectively, these sections provide a comprehensive understanding of how AI-driven recommendation systems (ADRS) enhance value cocreation behavior (VCB) within online health communities (OHCs) through patient knowledge acquisition and readiness.\u003c/p\u003e"},{"header":"2. Theoretical Framework and Hypothesis Development","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Value Cocreation in Online Health Communities (OHCs)\u003c/h2\u003e \u003cp\u003eValue cocreation (VCC) in online health communities (OHCs) refers to the dynamic process in which patients collaboratively generate mutual value by sharing knowledge, providing emotional support, and actively participating in peer interactions and technology-mediated engagements. Rather than serving as passive recipients of care or information, patients in OHCs take on active roles as coproducers of value, shaping both the content and the quality of community experiences(Ahuja and Shree, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eRecent research has suggested that AI-based recommendation systems and social assistance search bots can facilitate such cocreation by providing personalized guidance and information, enhancing patients\u0026rsquo; self-efficacy and health literacy (Chen \u0026amp; Zhang, 2024). Through interaction with AI, patients learn to interpret health information, share insights, and collaborate with others, thereby improving the collective community value.\u003c/p\u003e \u003cp\u003eAccordingly, we propose that AI systems such as search bots and recommendation systems can enhance patient knowledge acquisition and subsequently increase value cocreation behavior (VCB). The results of prior studies regarding VCB in online communities are presented in Table\u0026nbsp;1.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable I. Prior studies on VCB for online communities\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eReference (Year)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePurpose of Study\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTypes of VCB/Behavior\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTheories, Sample \u0026amp; Methodology\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eKey Integrated Variables\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThe impact of value co-creation on consumer citizenship behavior: Based on consumer perspective (Yang et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo examine VCB (value cocreation) in short-video/livestreaming platform context, and how engagement affects citizenship behavior\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eConsumer engagement behavior (participation) \u0026amp; Citizenship behavior\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eStructural Equation Modelling (survey of users of a short-video platform)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eConsumer engagement, perceived value, citizenship behavior\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAntecedents and consequences of value co-creation in online shopping (Sanaji et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo assess how online shopping experience influences VCB and how VCB influences repurchase intention/eWOM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eParticipation behaviour (CPB) \u0026amp; Citizenship behaviour (CCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePLS-SEM on 195 online-shopping customers (questionnaire survey)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eOnline shopping experience, CPB, CCB, repurchase intention, negative eWOM\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eA Study on Factors Affecting the Value Co-Creation Behavior in Sharing Economy Context (Zou and Shao, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo identify factors influencing customer VCB in sharing-economy platform users (e.g. Airbnb)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eValue co-creation behavior (engagement/citizenship)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSurvey of 587 sharing-economy platform users; structural modelling (SPSS/AMOS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCorporate (platform) authenticity, corporate image, ethical management; user characteristics; social capital; self-monitoring\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHow is brand love shaped through virtual event value co-creation and sponsorship: a multi-method approach (Xiao Fei et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo explore how participation and citizenship behaviours in virtual events affect \u0026ldquo;brand love\u0026rdquo;, includes VCB in virtual/online event settings\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eParticipation \u0026amp; Citizenship behaviour (as components of VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMulti-method (survey\u0026thinsp;+\u0026thinsp;qualitative) study analysing virtual event participants.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eVirtual event co-creation behaviours, sponsorship interactions, brand love/emotional attachment\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHow does the online innovation community climate affect the formation of users\u0026rsquo; value co-creation behavior (Tan et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo examine how community climate (supportive vs. controlling) influences user VCB in online innovation community\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eParticipation behavior \u0026amp; Citizenship behavior (user VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEmpirical analysis using data from a large online community (14 product-sections), behavioural data\u0026thinsp;+\u0026thinsp;modelling (Mplus)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCommunity climate (supportive, controlling), motivation (needs for achievement, power, affiliation), community trust, CPB, CCB\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCustomer citizenship behavior and customer perceived value in China: the mediating role of value co-creation experience (Yin et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo investigate how Customer Citizenship Behavior (CCB) leads to co-creation experience and perceived value in virtual brand communities\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCitizenship behavior (CCB)\u0026thinsp;+\u0026thinsp;cocreation experience as mediating process\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTwo-round data collection with total 642 matched questionnaires in virtual brand communities; structural model testing.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCCB dimensions, co-creation experience dimensions, customer perceived value (CPV)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Recommendation Systems and Search Bots in Online Communities\u003c/h2\u003e \u003cp\u003eRecommendation systems (RSs) and search bots are integral tools for enhancing the user experience within online communities by improving content discoverability and providing personalized recommendations. Recommendation systems optimize the presentation of content by using algorithms to suggest relevant posts, comments, or threads to users on the basis of their interests, preferences, and prior interactions. These systems are particularly useful in communities where users engage with large volumes of content, as they help prioritize the most relevant or helpful information. For example, in health-related forums, when users inquire about specific diseases, recommendation systems can surface the most pertinent responses, expert opinions, or similar discussions, enabling users to access information quickly and efficiently (Jannach and Adomavicius, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Ricci et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSearch bots, on the other hand, enhance the traditional search experience by interpreting user queries more effectively, showing the most relevant comments, threads, or posts at the top of search results. These bots often incorporate natural language processing (NLP) and machine learning techniques to better understand and process user input (Qiu et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), ensuring more accurate results. In the context of online communities, search bots can answer user queries instantly by retrieving and ranking content on the basis of various factors, such as textual relevance, user feedback (likes or upvotes), and engagement (Aggarwal, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; \u0026ldquo;(PDF) Collaborative Filtering Recommender Systems\u0026rdquo;, n.d.). The synergy between recommendation systems and search bots ensures that users not only find the right content but are also guided toward further exploration of related topics, fostering a richer, more interactive community experience(Jannach and Adomavicius, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Leskovec et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Qiu et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Ricci et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Trust in AI\u003c/h2\u003e \u003cp\u003eTrust in AI refers to users\u0026rsquo; belief in the competence, integrity, and benevolence of AI systems (Glikson and Woolley, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In the healthcare context, trust determines patients\u0026rsquo; willingness to rely on AI recommendations, affecting both acceptance and engagement (Febri Ramanda et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). When patients perceive an AI-driven system as transparent, consistent, and accurate, they are more likely to learn from its feedback and participate actively in online interactions. Conversely, low trust may inhibit knowledge exchange or sharing behaviors.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Research Framework and Hypotheses","content":"\u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Perceived Social Assistance from AI-Driven Recommendation Systems (ADRS) and Value Co-Creation Behavior (VCB)\u003c/h2\u003e \u003cp\u003eAI-driven recommendation systems (ADRSs) transform patient engagement in online health communities (OHCs) by offering personalized, algorithmically tailored support on the basis of user behavior, preferences, and health data. These systems enhance value cocreation behavior (VCB), a framework in which patients become active contributors to healthcare discussions and decision-making, by fulfilling social and informational needs that are central to social assistance theory (SAT) (Vickery et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Informational support refers to the delivery of personalized, health-related content, such as treatment options, lifestyle suggestions, and self-care guides, matching patient-specific needs and preferences. Recent studies have demonstrated that such AI-generated recommendations not only increase knowledge but also stimulate information-sharing behaviors (ISB) among patients in OHCs (Yang et al., \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Informational support plays a pivotal role in empowering users to make informed decisions, increasing their confidence in discussing health challenges within digital communities (Wei, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2025a\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn addition to facts and advice, ADRS increasingly incorporates emotional artificial intelligence (EAI) to simulate empathy through tone-sensitive messaging, sentiment-based prompts, and contextual reassurance. This fosters emotional connection and a sense of psychological safety, which encourages community citizenship behaviors such as feedback and support to peers (Miyazaki and Haderlie, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025a\u003c/span\u003e). Emotional support is a major predictor of sustained engagement in mental health-focused OHCs, especially among young adults seeking anonymity and empathy (Naga et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Companionship in ADRS is cultivated through continuous presence, community connectivity features, and AI-enabled nudges that recommend peer groups, shared experiences, and relevant conversations. These interactions reduce patient isolation, foster mutual encouragement, and promote a sustained sense of belonging, which strengthens participation in OHC dialogues(Latif et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eBy integrating IS, ES, and COMP, the ADRS drives both participation behaviors (e.g., information sharing, active browsing) and citizenship behaviors (e.g., peer support, feedback provision). These systems act as value enablers within OHCs, enhancing not only individual health literacy but also the collective intelligence and emotional resilience of the patient community. This layered support helps sustain engagement, reduces attrition, and supports long-term digital health transformation (Zhang and Lu, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTherefore, the study hypothesizes the following:\u003c/p\u003e \u003cp\u003eY1: Social assistance perceived (SAP) from AI-driven recommendation systems (ADRS) positively impacts patients' value cocreation behavior (VCB) in online health communities (OHCs).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Mediating Influence of Patient Knowledge Acquisition\u003c/h2\u003e \u003cp\u003eIn recent years, AI-driven recommendation systems (ADRS) have increasingly become foundational tools across online platforms, delivering not only personalized content but also critical social and emotional support to users. In OHCs and other digital environments, ADRS act as adaptive learning companions, enhancing user experiences by dynamically responding to individual information needs (Olawade et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Through informational support (IS), the ADRS provides customized, health-relevant content, such as lifestyle tips, symptom explanations, and medication guidance, thereby strengthening patients\u0026rsquo; knowledge acquisition capacity and autonomy. Recent studies in digital education contexts also show how AI agents significantly enrich learning by delivering real-time, relevant feedback and curated resources tailored to user needs (Vanderhout et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThese systems operate continuously, enabling users to access health knowledge or ask questions at any time. Thus, fostering ongoing, autonomous learning (Miyazaki and Haderlie, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025a\u003c/span\u003e). Moreover, the emotional support (ES) offered by ADRS plays a key role in reinforcing users\u0026rsquo; commitment to learning. Empathetic responses, motivational cues, and contextual affirmations help maintain patients' focus and perseverance, particularly in long-term health management scenarios (Naga et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). ADRS further serve as personalized companions by adapting their guidance to fit each user\u0026rsquo;s behavioral and emotional patterns, creating engaging and individualized learning experiences (Latif et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In OHCs, this facilitates the delivery of digestible and conversational health information, empowering patients to engage in interactive, intuitive learning and improving their capacity for health-related decision-making.\u003c/p\u003e \u003cp\u003eY2a: Social assistance perceived (SAP) from AI-driven recommendation systems (ADRS) has a positive effect on patient knowledge acquisition (PKA) in OHCs.\u003c/p\u003e \u003cp\u003eResearch on digital communities emphasizes the importance of customer learning, where individuals develop expertise through content consumption, discussions, and collaboration. This learning empowers users to become knowledge integrators who contribute value back into the community through engagement and cocreation behaviors(Vickery et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In OHCs, patient knowledge acquisition (PKA) involves absorbing health knowledge and applying it to personal or shared experiences, which encourages participation in discussions, peer mentoring, and treatment feedback activities (Zhang and Lu, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eY2b: Patient knowledge acquisition (PKA) has a constructive effect on patients\u0026rsquo; value cocreation behavior (VCB) in OHCs.\u003c/p\u003e \u003cp\u003eADRS streamlines the learning process by minimizing effort and maximizing relevance through real-time interaction and adaptive content filtering. The social interaction mechanisms embedded within ADRS have been shown to improve users' cognitive engagement, enhance collective learning, and drive behavioral participation (Miyazaki and Haderlie, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025a\u003c/span\u003e). When integrated into OHCs, these features support immersive learning environments, where patients not only acquire knowledge but also engage in collaborative learning loops such as asking questions, offering advice, and sharing experiences with others (Vanderhout et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). This continuous exchange is further strengthened by companionship (COMP) elements of ADRS, which connect patients with similar conditions, reinforcing peer-to-peer learning and emotional bonding (Olawade et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Emotional reinforcement within these platforms encourages a sense of accountability and prosocial behaviors, such as giving feedback, offering reassurance, or initiating support threads (Latif et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Despite increasing attention given to PKA in digital health spaces, the specific mechanism by which PKA mediates the link between perceived support and cocreation behaviors remains underexplored.\u003c/p\u003e \u003cp\u003eY2c: Patient knowledge acquisition (PKA) mediates the relationship between social assistance perceived (SAP) from the ADRS and patients\u0026rsquo; value cocreation behavior (VCB) in OHCs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3. The Moderating Role of Patient Ability/Readiness\u003c/h2\u003e \u003cp\u003ePatient ability/readiness (PAR) refers to an individual\u0026rsquo;s cognitive, emotional, and technical preparedness to learn and adopt new systems or knowledge. In digital health contexts, it encompasses patients\u0026rsquo; confidence and ability to interact with AI systems, understand information, and apply it in meaningful ways (Ramadan et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In particular, within AI-driven recommendation systems (ADRS), PAR significantly shapes how patients perceive and utilize the support provided. For example, patients with greater readiness are better equipped to engage with health information and navigate complex AI interfaces. Conversely, those with limited readiness may require simplified interfaces, structured guidance, and more emotional reinforcement to derive value from ADRS (Ain et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGiven that the ADRS functions as both an educational tool and a support tool in online health communities (OHCs), the impact of social assistance perceived (SAP) on patient knowledge acquisition (PKA) is expected to vary depending on the patient\u0026rsquo;s level of readiness. High-PAR individuals are more likely to independently engage with IS and seek out ADRS for assistance, whereas low-PAR users may require motivational or emotional nudges to fully participate (Vinh and Hung, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Therefore:\u003c/p\u003e \u003cp\u003eY3: Patient ability/readiness (PAR) positively moderates the relationship between social assistance perceived (SAP) from the ADRS and patient knowledge acquisition (PKA) in OHCs.\u003c/p\u003e \u003cp\u003ePAR has been recognized across domains as a predictor of digital tool adoption and behavioral intent. Studies on AI integration suggest that users with low digital confidence often disengage or exhibit avoidance behavior, especially when the systems are perceived as complex or ambiguous (Ain et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In contrast, users with high readiness demonstrate stronger adoption motivation, learning persistence, and peer collaboration, leading to richer interaction with AI systems (Ramadan et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In the context of OHCs, the importance of incorporating PAR stems from the varying levels of digital literacy and self-efficacy among patients. Those who possess the ability to interact effectively with ADRS are more likely to leverage SAP for deeper learning and value cocreation behavior (VCB). In contrast, those with lower readiness may encounter cognitive or emotional barriers, reducing their ability to participate in and benefit from community engagement activities (Vinh and Hung, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). This underscores the need to consider PAR in a moderated mediation framework, where it not only influences the direct relationship between the SAP and PKA but also affects the strength of the indirect path from the SAP to the VCB via the PKA. In such a framework, we anticipate the following:\u003c/p\u003e \u003cp\u003eY4: The indirect effect of SAP from ADRS on patients' VCB through the PKA will be stronger when patient ability/readiness (PAR) is high.\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, this moderated mediation model captures the conditional nature of ADRS efficacy in OHCs. It posits that ADRS can enhance patient outcomes across the board, but their impact will be magnified for those with greater readiness, who are better positioned to engage, learn, and contribute. On the other hand, individuals with lower readiness may experience attenuated benefits, necessitating adaptive system designs that account for digital literacy variability and learning support needs. A summary of the above results is presented in Table\u0026nbsp;2.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable II. Summary of Constructs and Hypotheses\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabb\" border=\"1\"\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConstruct\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDefinition\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHypothesis\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eExpected Relationship\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial Assistance Perceived (SAP)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSupport provided by AI-driven recommendation systems (ADRS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eY1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+ on PKA\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Knowledge Acquisition (PKA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThe process by which patients acquire knowledge from ADRS and peers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eY2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+ on VCB\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Ability/Readiness (PAR)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThe capacity and readiness of patients to interact with ADRS and engage\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eY3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eModerates the effect of SAP on PKA\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Knowledge Acquisition (PKA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMediator between SAP and VCB in OHCs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eY4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMediates the relationship between SAP and VCB\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Methodology","content":"\u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Research Design\u003c/h2\u003e \u003cp\u003eThis study employs a quantitative, cross-sectional approach to empirically evaluate the theoretical framework addressing AI-enabled value cocreation within online health communities (OHCs). Data were collected via a structured online questionnaire administered to patients who are active participants in OHCs that incorporate AI-based tools such as recommendation systems, conversational agents, or personalized digital health assistants. For data analysis, this study uses partial least squares structural equation modelling (PLS-SEM) to validate the proposed model and multilayer perceptron (MLP) techniques to investigate nonlinear relationships and assess the relative importance of predictors. The integration of these methods is in line with current best practices for analysing complex models in technology adoption research (Ketchen, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Data Collection\u003c/h2\u003e \u003cp\u003eData for this study were collected from active members of online health communities (OHCs) in China, with a focus on both patient\u0026ndash;to-patient and patient\u0026ndash;doctor interactions. To ensure contextual relevance, participants were presented with examples of six well-known Chinese OHC platforms, each hosting between approximately 450,000 and 3.8\u0026nbsp;million registered members, demonstrating high levels of digital engagement. The study specifically targeted individuals who had experience using AI-driven recommendation systems (ADRS) embedded within these OHCs, as these systems assist users by offering tailored health recommendations, emotional guidance, and informational support. Within these online communities, members frequently engage in collaborative health-related behaviors such as exchanging medical advice, sharing treatment experiences, offering feedback on healthcare services, and supporting other patients\u0026rsquo; recovery processes. Given their interactive and participatory nature, OHCs provide an ideal environment for examining value cocreation behavior (VCB).\u003c/p\u003e \u003cp\u003e Prior to data collection, ethical approval was granted by the university\u0026rsquo;s Institutional Ethics Review Committee. The study adhered strictly to ethical research standards by ensuring participant confidentiality, voluntary participation, and informed consent. All respondents were informed that the collected data would be used solely for academic purposes and that anonymity would be preserved. The data for this study was collected using random sampling, a probability sampling technique where participants are selected at random from the population. This approach ensures that every individual has an equal chance of being included in the sample, which helps to reduce bias and increase the generalizability of the results.\u003c/p\u003e \u003cp\u003eTwo screening questions were used to ensure sample validity. First, participants were asked to identify the OHCs they actively used to verify their familiarity with such platforms. Second, they were asked to confirm whether they had recently interacted with ADRS tools within these communities. Only respondents meeting both criteria were permitted to complete the main questionnaire, which was administered online. The survey, designed to be completed within 10\u0026ndash;12 minutes, was distributed over a three-month period from March 2025 to May 2025. Out of the 780 distributed questionnaires, 476 responses were received. After incomplete and inconsistent responses, were excluded, 362 valid responses were retained for the final analysis. The demographic profile of the respondents is presented in Table\u0026nbsp;3, and our expected data characteristics are presentd in Table\u0026nbsp;4.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eTable III. Sample characteristics\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabc\" border=\"1\"\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDemographics Items\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFrequency\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePercentage (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e57.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e162\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e43.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e21\u0026ndash;30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e131\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e38.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e31\u0026ndash;40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e41\u0026ndash;50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e51\u0026ndash;60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e61 or above\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSchooling Level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh school or lower\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBachelor\u0026rsquo;s degree\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e235\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e67.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaster\u0026rsquo;s degree or higher\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTenure of using OHC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;6 months\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e198\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u0026ndash;12 months\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e13\u0026ndash;36 months\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;37 months\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRecent experience with OHC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;7 days\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7 days to \u0026lt;\u0026thinsp;15 days\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e130\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e38.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e15 days to \u0026lt;\u0026thinsp;1 month\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1 month to \u0026lt;\u0026thinsp;2 months\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eTable IV. Expected Data Characteristics\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabd\" border=\"1\"\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatistic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eExpected Range\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNotes\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean age\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e25\u0026ndash;45 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eActive digital health users\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e~\u0026thinsp;55% female\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCommon in OHC participation\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOHC experience\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6\u0026ndash;24 months\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eModerate engagement\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eReliability (α)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.80\u0026ndash;0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIndicates high internal consistency\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eR\u0026sup2; (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.45\u0026ndash;0.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSubstantial explanatory power\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Construct Measurement\u003c/h2\u003e \u003cp\u003eThe measures used in this study were adapted from existing scales in prior research to ensure consistency with established methodologies (Liu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Yi and Gong, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). To assess each construct, a 7-point Likert scale was used, where 1 indicated \u0026ldquo;strongly disagree\u0026rdquo; and 7 indicated \u0026ldquo;strongly agree.\u0026rdquo; The items for each construct were slightly modified to suit the specific context of this study, particularly in relation to online health communities (OHCs) and AI-driven recommendation systems (ADRS). Since the research was conducted in China, the questionnaire followed a double translation protocol (Harkness et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Initially, the English version was translated into Chinese, and then translated back into English by two bilingual experts from the management faculty, who were familiar with the technical terminology but were unaware of the study's aims. The back-translated versions did not reveal any significant differences from the original versions.\u003c/p\u003e \u003cp\u003eThe dependent variable for this study, patients' value cocreation behavior (VCB), was measured via items adapted from (Yi and Gong, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) and (Liu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). VCB can be divided into two main categories: participation behavior and citizenship behavior. Participation behavior was further subdivided into information sharing behavior (ISB) and responsible behavior (RB), with ISB including 5 items and RB consisting of 4 items. Citizenship behavior was measured via Advocacy Behavior (AB) and Feedback Behavior (FB), with each containing 3 items. The relationships among participation behavior, citizenship behavior, and VCB can be expressed as a weighted sum, where VCB is a function of participation and citizenship behaviors. The formula for VCB is Eq.\u0026nbsp;1.\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:VCB=\\alpha\\:\\cdot\\:PB+\\beta\\:\\cdot\\:CB\\)\u003c/span\u003e \u003c/span\u003e \u003cem\u003eEq.\u0026nbsp;1\u003c/em\u003e\u003c/p\u003e \u003cp\u003ewhere PB is the total number of participation behaviors (ISB\u0026thinsp;+\u0026thinsp;RB), CB is the total number of citizenship behaviors (AB\u0026thinsp;+\u0026thinsp;FB), and α and β represent the weights of each behavior type on the basis of their influence on VCB. For the independent variable, social assistance perceived (SAP) from AI-driven recommendation systems (ADRS) in OHCs, the construct was operationalized as a multidimensional variable, following (Liu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The SAP is broken down into three key dimensions: informational support (IS), which consists of 3 items; emotional support (ES), which comprises 4 items; and companionship (COMP), which has 2 items. The overall SAP construct was calculated as the sum of these three dimensions, weighted by their respective importance in the context of ADRS. This can be represented as Eq.\u0026nbsp;2.\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:PSS=\\sum\\:({\\omega\\:}_{i}\\cdot\\:I{S}_{i})+\\sum\\:({\\theta\\:}_{j}\\cdot\\:E{S}_{j})+\\sum\\:({\\varphi\\:}_{k}\\cdot\\:COM{P}_{k})\\)\u003c/span\u003e \u003c/span\u003e \u003cem\u003eEq.\u0026nbsp;2\u003c/em\u003e\u003c/p\u003e \u003cp\u003ewhere ω, θ, and φ represent the weights for each dimension, and IS, ES, and COMP refer to informational support, emotional support, and companionship, respectively. The mediator variable, patient knowledge acquisition (PKA), was adapted from (Zhao, \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and consists of 3 items tailored for the OHC context. The PKA represents the process through which patients acquire knowledge from interactions within these communities, particularly those mediated by. The relationship between SAP and PKA was modelled via a linear regression approach, where PKA is predicted by the social assistance perceived (SAP):\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:PKA={\\delta\\:}_{1}\\cdot\\:SAP+ϵ\\)\u003c/span\u003e \u003c/span\u003e Eq.\u0026nbsp;3\u003c/p\u003e \u003cp\u003ewhere δ1 is the coefficient representing the effect of the SAP on the PKA, and where ε is the error term. For the moderator variable, patient ability/readiness (PAR), the scale was adapted from (Gao et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) with 4 items adjusted for the OHC context. PAR captures the readiness and digital competence of patients to engage with ADRS. The moderating effect of PAR on the relationship between SAP and PKA was modelled via an interaction term, which captures how PAR influences the effect of SAP on PKA. This relationship is expressed as:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:PKA={\\delta\\:}_{2}\\cdot\\:SAP+\\gamma\\:\\cdot\\:PAR+\\eta\\:\\cdot\\:(SAP\\times\\:PAR)+ϵ\\)\u003c/span\u003e \u003c/span\u003e Eq.\u0026nbsp;4\u003c/p\u003e \u003cp\u003ewhere δ2, γ, and η are the coefficients for each term, representing the direct effects of SAP, PAR, and the interaction between SAP and PAR, respectively. Finally, the study includes five control variables: gender, age, education, recent experience with OHCs, and tenure of use (Bu \u003cem\u003eet al\u003c/em\u003e., 2022). These control variables were included to account for potential confounding factors, allowing for a more accurate analysis of the primary constructs. The impact of these control variables on VCB was modelled as part of a multiple regression analysis:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:VCB={\\beta\\:}_{1}\\cdot\\:SAP+{\\beta\\:}_{2}\\cdot\\:PKA+{\\beta\\:}_{3}\\cdot\\:PAR+\\sum\\:({\\theta\\:}_{n}\\cdot\\:Contro{l}_{n})+ϵ\\)\u003c/span\u003e \u003c/span\u003e Eq.\u0026nbsp;5\u003c/p\u003e \u003cp\u003ewhere Control_n represents the control variables (e.g., gender, age), and β1, β2, β3, and θn are the coefficients for each respective term. This model allows the effects of the primary constructs to be isolated while accounting for the influence of demographic and contextual factors.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Data Analysis and Validation","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Data Analysis\u003c/h2\u003e \u003cp\u003eThe data were analysed in two phases. First, we used SmartPLS 4.0 to perform partial least squares structural equation modelling (PLS-SEM) (Ringle \u003cem\u003eet al\u003c/em\u003e., 2021). This method has gained significant traction across various disciplines because of its ability to handle complex models and data with nonnormal distributions, which are often encountered in social science research (Sarstedt \u003cem\u003eet al\u003c/em\u003e., 2021). PLS-SEM is especially valuable when relationships between exogenous and endogenous variables are nonlinear. Linear models can fail to capture the intricacies inherent in phenomena such as value cocreation behavior (VCB). To address this issue, we integrated a neural network approach, which is capable of modelling both linear and nonlinear relationships among constructs. Scholars have increasingly advocated the use of hybrid approaches that combine PLS-SEM with machine learning techniques to increase model robustness and generalizability (Hair \u003cem\u003eet al\u003c/em\u003e., 2020). Furthermore, to test the reliability and validity of the study model, we utilized a multilayer perceptron (MLP), following the recommendations of recent advancements in methodology (Shahzad \u003cem\u003eet al\u003c/em\u003e., 2021). This dual-approach methodology provided a more comprehensive understanding of the relationships among the key variables in our model.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable V. Data Analysis Summary\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabe\" border=\"1\"\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStage\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMethod\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePurpose\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOutput\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePLS-SEM(SmartPLS 4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTest hypotheses, mediation, moderation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePath coefficients, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e, f\u0026sup2;, VIF\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMLP (IBM SPSS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIdentify variable importance, capture nonlinearity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRMSE, R\u0026sup2;, Normalized Importance\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBootstrapping\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eConfirm robustness\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eConfidence Intervals (CI), significance\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Measurement Validation\u003c/h2\u003e \u003cp\u003eTo assess the measurement model and test the hypotheses, we employed SmartPLS 4 (Ringle et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Following the guidelines of (Latan et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), a series of validity and reliability tests were performed. The internal consistency of the constructs was validated via Cronbach\u0026rsquo;s alpha (α) and composite reliability (CR), both of which exceeded the recommended threshold of 0.70, indicating that the constructs are both reliable and consistent. Additionally, the average variance extracted (AVE) was calculated, with all values exceeding the minimum threshold of 0.50, confirming adequate convergent validity (Hair and Alamer, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The results for α, CR, AVE, and factor loadings are summarized in Table\u0026nbsp;6.\u003c/p\u003e \u003cp\u003eGiven that AI-driven recommendation systems (ADRS) and value cocreation behavior (VCB) are modelled as second-order formative constructs, traditional reliability and validity tests are not suitable because of their directional causality and expected lower correlations among formative indicators, which should not be disregarded (Hair et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). As per (Hair et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), we assessed the outer weights for the ADRS and VCB dimensions, ensuring that they met the significance threshold of p\u0026thinsp;\u0026lt;\u0026thinsp;0.001. Furthermore, multicollinearity tests were conducted for the second-order formative constructs, and the variance inflation factor (VIF) values for the measured items were all below the threshold of 5.0, confirming that multicollinearity issues were not present. These results, presented in Table\u0026nbsp;7, indicate that the second-order formative constructs exhibit strong validity, and no multicollinearity concerns were identified.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable VI. Reliability and Validity of the Constructs\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabf\" border=\"1\"\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConstruct\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eItems\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLoadings\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eα\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCR\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial Assistance Perceived (SAP) from ADRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.918\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.912\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.586\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInformational Support (IS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIS1, IS2, IS3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.885\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.668\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEmotional Support (ES)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eES1, ES2, ES3, ES4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.824\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.899\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.742\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCompanionship (COMP)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCOMP1, COMP2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.795\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.911\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.829\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Knowledge Acquisition (PKA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePKA1, PKA2, PKA3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.764\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.872\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.678\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Ability/Readiness (PAR)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePAR1, PAR2, PAR3, PAR4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.836\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.674\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValue Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.935\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.939\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.523\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInformation Sharing Behavior (ISB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eISB1, ISB2, ISB3, ISB4, ISB5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.885\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.912\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.682\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResponsible Behavior (RB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRB1, RB2, RB3, RB4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.818\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.735\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFeedback Behavior (FB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFB1, FB2, FB3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.798\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.883\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.718\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eTable VII. Evaluation of Second-Order Formative Constructs with VIF Values for Latent Variables\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabg\" border=\"1\"\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConstructs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWeight Coefficients\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eT-Statistics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eP Values\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eVIF (Variance Inflation Factor)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValue Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.253\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e23.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.27\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValue Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.229\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.375\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValue Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.422\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e30.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3.63\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValue Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.245\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e25.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.31\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAI-Driven Recommendation Systems (ADRS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.505\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e31.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.235\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAI-Driven Recommendation Systems (ADRS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.358\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAI-Driven Recommendation Systems (ADRS)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.269\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e23.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.88\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e5.3 Common Method Bias (CMB)\u003c/h2\u003e \u003cp\u003eCommon method bias (CMB) remains a concern in survey-based research, particularly when data are collected from a single source (Podsakoff et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). To address this issue, we employed two well-established techniques. First, we conducted Harman\u0026rsquo;s single-factor test (Kock, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), which revealed that no single factor explained more than 38% of the variance, well below the conventional threshold of 50%. This finding indicates that CMB is not a significant issue in our study. The second method involved performing a full collinearity assessment via SmartPLS 4, following the approach suggested by (Kock, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). This technique, which is gaining widespread acceptance in contemporary research, assesses multicollinearity and ensures the validity of the structural model. The variance inflation factor (VIF) values reported in Table\u0026nbsp;8 indicate that most values are well below the threshold of 3.3. Only one value, ISB (6) (VIF\u0026thinsp;=\u0026thinsp;3.369), exceeded this threshold, but it remains within the acceptable range of 5, as recommended by (Kock, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) Both methods confirm that CMB is not a significant concern in our study.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable VIII. Full Collinearity Test of First-Order Constructs on VCB\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabh\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConstructs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVIF\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdvocacy Behavior (AB) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.342\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCompanionship (COMP) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.315\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEmotional Support (ES) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.731\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFeedback Behavior (FB) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.305\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInformational Support (IS) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.149\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInformation Sharing Behavior (ISB) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3.42\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient knowledge Acquisition (PKA) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.89\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Ability/Readiness (PAR) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.426\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResponsible Behavior (RB) \u0026rarr; Value Co-Creation Behavior (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.318\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e5.4. Discriminant Validity\u003c/h2\u003e \u003cp\u003eWe applied two widely recognized methods to assess discriminant validity. The first approach used to evaluate discriminant validity in this study was Fornell and Larcker\u0026rsquo;s criterion (1981), which posits that the square roots of the average variance extracted (AVE) for each construct should exceed the correlations between the constructs. The second method applied was the heterotrait-monotrait (HTMT) ratio of correlations, as suggested by (Henseler et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). According to this more recent method, the HTMT value should remain below the threshold of 0.85. The results obtained from both techniques in this study satisfy the necessary criteria. Additional details are provided in Tables\u0026nbsp;9 and 10.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable IX. Fornell-Larcker Criterion\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabi\" border=\"1\"\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConstruct\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eC1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eC2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eC3\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eC4\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eC5\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eC6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eC7\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eC8\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.867\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.456\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.574\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.648\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.672\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.586\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCOMP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.466\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.916\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.624\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.682\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.547\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.645\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.655\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eES\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.563\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.665\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.874\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.609\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.773\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.671\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.698\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.649\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.503\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.521\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.836\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.633\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.596\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.494\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.612\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.687\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.546\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.857\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.532\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.623\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eISB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.675\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.751\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.689\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.532\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.831\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.691\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.763\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePKA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.634\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.688\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.696\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.607\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.612\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.719\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.834\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.625\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.602\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.558\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.539\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.694\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.732\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.845\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.615\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.562\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.633\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.563\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.703\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.624\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.849\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eTable X. Heterotrait-Monotrait Ratio (HTMT)\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabj\" border=\"1\"\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConstruct\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eC1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eC2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eC3\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eC4\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eC5\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eC6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eC7\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eC8\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.626\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.704\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.749\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.633\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.688\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.628\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCOMP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.578\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.812\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.813\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.659\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.693\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.818\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.785\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.692\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eES\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.659\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.787\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.708\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.802\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.77\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.633\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.653\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.692\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.782\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.713\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.732\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.792\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.721\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.819\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.781\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eISB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.763\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.753\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.702\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.812\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.749\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePKA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.802\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.824\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.797\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.759\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.819\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.705\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.681\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.658\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.622\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.808\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.831\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.735\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.686\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.654\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.801\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.694\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.818\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.823\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.761\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe relationships between the underlying variables were examined via R\u0026sup2; values. According to (Hair et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), R\u0026sup2; values of 0.25, 0.50, and 0.75 correspond to weak, moderate, and substantial explanatory power, respectively. In our model, the R\u0026sup2; values were 0.552, 0.540, and 0.633, indicating moderate to substantial levels of explanation. Furthermore, the structural model results met the goodness-of-fit criteria, with RMSEA\u0026thinsp;=\u0026thinsp;0.053 (below the threshold of 0.080) and SRMR\u0026thinsp;=\u0026thinsp;0.073 (also under the acceptable threshold of 0.080). Thus, all model fit criteria were satisfied (Hair et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). To verify the results, we performed a bootstrap procedure with 5,000 samples to test the significance of all paths in the structural model (Hair and Alamer, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The results of the hypothesis tests for this study are shown in Table\u0026nbsp;11, and they were analysed in three stages.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable XI. Results of Hypothesis Testing\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabk\" border=\"1\"\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eModel 1: β\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eModel 1: T-statistics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eModel 2: β\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eModel 2: T-statistics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eModel 3: β and T-statistics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eR\u0026sup2;\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(ADRS) to (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.733\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9.811\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.433\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5.702\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.428\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.552\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(ADRS) to Patient Knowledge Acquisition (PKA)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.822\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11.123\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.315\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.512\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.540\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Knowledge Acquisition (PKA) to (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.385\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5.213\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.385\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5.213\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.633\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(ADRS) \u0026rarr; Patient Knowledge Acquisition to VCB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.315\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4.671\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(PAR) to (VCB)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e-0.185\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2.492\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePAR \u0026times; ADRS to PKA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.213\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5.023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn Model 1, we initially tested the direct impact of AI-driven recommendation systems (ADRS) on value cocreation behavior (VCB). The results indicated a significant positive relationship between ADRS and VCB (β\u0026thinsp;=\u0026thinsp;0.733, t\u0026thinsp;=\u0026thinsp;9.811, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), providing strong support for H1. Next, in Model 2, we explored a mediation model based on the framework introduced by (Latif et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), which considers both direct and indirect pathways. The results demonstrated that the ADRS positively influences patient knowledge acquisition (PKA) (β\u0026thinsp;=\u0026thinsp;0.822, t\u0026thinsp;=\u0026thinsp;11.123, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and PKA, in turn, significantly impacts VCB (β\u0026thinsp;=\u0026thinsp;0.385, t\u0026thinsp;=\u0026thinsp;5.213, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). These findings support H2a and H2b. The indirect effect in Model 2 (β\u0026thinsp;=\u0026thinsp;0.315, t\u0026thinsp;=\u0026thinsp;4.671, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) further confirms the mediation process. The reduction in effect size when PKA is included in the model suggests partial mediation, confirming H2c.\u003c/p\u003e \u003cp\u003eIn Model 3, we examined the moderating role of Patient Ability/Readiness (PAR). Initially, we tested the direct effect of PAR on the relationship between ADRS and VCB (β = -0.185, t\u0026thinsp;=\u0026thinsp;2.492, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) without considering the interaction effect. Next, we evaluated the interaction effect of ADRS and PAR on VCB, and the results indicated that PAR significantly moderated the relationship between ADRS and PKA (β\u0026thinsp;=\u0026thinsp;0.213, t\u0026thinsp;=\u0026thinsp;5.023, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), thus supporting H3. The significant interaction slopes are visually represented in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Finally, we conducted a moderated mediation analysis via bias-corrected bootstrapping, as outlined by Preacher \u003cem\u003eet al\u003c/em\u003e. (2023). The results presented in Table\u0026nbsp;11 show that the indirect effect of the ADRS on VCB through the PKA varies at different PARs. Specifically, when the PAR is high, the indirect effect of ADRS on VCB is significant (β\u0026thinsp;=\u0026thinsp;0.186, 95% bias-corrected CI [0.103, 0.269]), whereas when the PAR is low, this effect becomes nonsignificant (β\u0026thinsp;=\u0026thinsp;0.038, boot SE\u0026thinsp;=\u0026thinsp;0.095, 95% bias-corrected CI [-0.067, 0.136]). Therefore, the positive indirect effect of the ADRS on VCB through the PKA is present when the PAR is high, supporting H4. The control variables in the study, including gender (β = -0.041), age (β\u0026thinsp;=\u0026thinsp;0.032), education (β\u0026thinsp;=\u0026thinsp;0.011), recent experience with OHCs (β\u0026thinsp;=\u0026thinsp;0.015), and tenure of OHC use (β\u0026thinsp;=\u0026thinsp;0.012), were found to be insignificant.\u003c/p\u003e \u003c/div\u003e"},{"header":"6. Robustness Checks: Multilayer Perceptron (MLP) Approach","content":"\u003cp\u003eTo validate the proposed research model and examine the hypotheses, this study employs a hybrid analytical approach that integrates structural equation modelling (SEM) and a multilayer perceptron (MLP). Initially, PLS-SEM was utilized to assess the reliability and validity of the measurement model, with a primary focus on evaluating the constructs and their indicators. Afterward, the structural model was analysed to verify the hypothesized relationships among the constructs. Subsequently, MLP analysis was performed to increase the model's accuracy, further identifying the critical factors that influence value cocreation behavior (VCB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eTable XII. Conditional Indirect Effect of the ADRS on VCB, through the PKA at Different PAR Values\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabl\" border=\"1\"\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBoot Effect\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLLCI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eULCI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePAR Values\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e+\u0026thinsp;1 SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.192\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.109\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.276\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e-1 SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.047\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-0.072\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.205\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eCompared with conventional regression methods, multilayer perceptron (MLP) methods are recognized for their superior predictive power, offering more accurate forecasts (Kadir et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Schroer and Just, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Previous studies suggest that integrating structural equation modelling (SEM) with multilayer perceptron (MLP) techniques can significantly advance the field of information systems, potentially establishing a new paradigm for research methodologies (Leong et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn this study, we applied the widely used multilayer perceptron (MLP) method, a common artificial intelligence approach, to train neural networks (Drałus et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The neural network model was built via SPSS 21.0, which incorporates multiple hierarchical layers. Typically, an MLP consists of an input layer, one or more hidden layers, and an output layer. While no standardized method exists for determining the optimal values, the configuration of the hidden layers is dependent on the complexity of the problem being addressed (Trigka and Dritsas, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The significance of the predictor variables was evaluated in two stages. Initially, three key covariates were introduced in the input layer, the SAP from the ADRS, the PKA, and the PAR, with VCB placed in the output layer. The sigmoid function was used as the activation function in both the hidden and output layers, following the methodology outlined by (Tao et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003cp\u003eAs recommended by (Tao et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and (Drałus et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), the MLP model was validated by testing with hidden nodes ranging from 1\u0026ndash;10. To mitigate overfitting, we employed tenfold cross-validation, with 70% of the data used for training and the remaining 30% for testing. The root mean square error (RMSE) was calculated to assess errors during both the training and testing phases, as shown in Table\u0026nbsp;13. The findings revealed that the average RMSE for value cocreation behavior (VCB) was 0.370 for the training set and 0.363 for the testing set, indicating minimal variation. These low values suggest a high degree of accuracy in predicting VCB, the endogenous construct. The detailed results are presented in Table\u0026nbsp;13.\u003c/p\u003e \u003cp\u003eFurthermore, a sensitivity analysis was performed to assess the contribution and importance of the covariates. The model's importance and normalized importance were evaluated by averaging the generated values of the constructs over ten iterations to predict the output. Normalized importance refers to the proportion of each input relative to the highest value, highlighting that patient ability/readiness (PAR) was the most influential predictor of VCB (importance value\u0026thinsp;=\u0026thinsp;0.422), followed by SAP from ADRS (0.307) and PKA (see Table\u0026nbsp;14). This underscores the importance of optimizing the most influential predictor to increase the effectiveness and overall performance of the model. While some minor adjustments in the model ranking were observed, the PAR and SAP from the ADRS maintained similar rankings throughout.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable XIII. Neural network validation for training and testing data\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabm\" border=\"1\"\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeural Networks\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraining Data N\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTraining Data RMSE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTesting Data N\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTesting Data RMSE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSum of Square Error\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e230\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.395\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.317\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e37.832\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e230\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.372\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e33.409\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e234\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.338\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e104\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.327\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e25.297\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e247\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.357\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.345\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e32.678\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e227\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.421\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.412\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e41.213\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e237\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.346\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.392\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e26.045\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e238\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.386\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.315\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e36.132\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e242\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.342\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.367\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e29.812\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e237\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.359\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.353\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e29.431\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e244\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.345\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.398\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e28.914\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.370\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"7. Discussion","content":"\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e7.1. Summary of Key Findings\u003c/h2\u003e \u003cp\u003eThis study aimed to develop and test a moderated mediation model explaining the formation of patients' value cocreation behavior (VCB) in online health communities (OHCs). The results indicated that the support provided by AI recommender systems (AIRS) positively influences patients' VCB, which is consistent with prior research in various ways, despite the relatively underexplored context and the novel nature of our model. Specifically, similar to the findings of (Li et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), social assistance plays a crucial role in influencing individuals' willingness to engage in community-based activities in digital commerce settings. However, our study uniquely focused on the specific contributions of various types of social assistance in shaping patient behavior toward value cocreation.\u003c/p\u003e \u003cp\u003eAdditionally, our findings align with the results of (Liu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), who demonstrated that social assistance is multidimensional and affects cocreation intentions. Unlike previous studies, this research incorporates multiple dimensions of social assistance, such as informational support (IS), emotional support (ES), and companionship (COMP), all of which significantly influence patients' VCB. We conceptualized SAP from ADRS as a second-order formative construct, as their integration provides a more holistic understanding of the latent variable. Furthermore, the findings expand upon the work of (Latif \u003cem\u003eet al.\u003c/em\u003e, 2025), reinforcing the idea that social assistance is an essential factor in the development of VCB.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTable XIV. Sensitivity Analysis\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabn\" border=\"1\"\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeural Networks\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSAP from ADRS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePAR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePKA\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.391\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.379\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.118\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.591\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.296\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.421\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.225\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.368\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.22\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.359\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.34\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.312\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.502\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.209\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.261\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.218\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.455\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.275\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.261\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.135\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.359\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.253\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.396\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.348\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage Importance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.307\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.422\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.271\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRelative Importance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.726\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.643\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNormalized Importance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e72.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e64.32\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThis study highlights the influential role of AI-based social assistance recommender systems (ADRSs) in inducing members' value cocreation behavior (VCB) in online health communities (OHCs), which is a traditionally human-driven process. Our research suggests that ADRS are more effective in fostering patient VCB within OHCs because of their continuous support, which helps patients remain engaged with others. For example, a search bot providing reliable informational support (IS) and companionship (COMP) encourages patients to interact confidently and contribute to community discussions. Recent research by (Li and Tuunanen, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) explores how social interactions and resource integration contribute to information technology-based value cocreation in service systems. This study further demonstrated that AI tools, such as ADRS, enhance patients' VCB. Unlike (Latif \u003cem\u003eet al.\u003c/em\u003e, 2025), who focused on the role of social assistance search bots in addressing global health challenges, our study specifically investigates how various types of social assistance, as perceived by recommenders and search bots, influence patient participation and contribution behaviors in OHCs to improve individual and collective health outcomes.\u003c/p\u003e \u003cp\u003eThe results of this study align with those of previous studies indicating that learning stems from multiple types of social assistance (Zhao, \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). For example, IS provided by search bots is instrumental in helping members learn in educational communities (Guan et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).Our findings indicate that ADRS functions as an antecedent to learning by providing patients access to relevant information and empathetic responses to their concerns, thereby enhancing their overall experience and engagement. Consistent with (Nohutlu et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), who emphasized that learning is a key motivator for customer value cocreation in online innovation communities, our study shows that patient knowledge acquisition (PKA) positively impacts VCB in OHCs. The positive influence of these variables suggests that patients in OHCs acquire knowledge and specific skills to address their health issues, which ultimately drives feedback behavior (FB). This aligns with the work of (Latif \u003cem\u003eet al.\u003c/em\u003e, 2025) and (Chen et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), confirming that learning serves as a mediator to explain how patients use ADRS to generate various VCBs. Unlike earlier studies by (Liu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) and (Latif \u003cem\u003eet al.\u003c/em\u003e, 2025), which examined member belongingness and ethical perception as mediators between social assistance and VCB, our study revealed that, in a supportive learning environment, patients feel more empowered and positive about participating in cocreation activities. Consequently, learning emerges as a crucial component of the sustainability of OHCs (Chen et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAdditionally, our findings show that patient ability/readiness (PAR) positively moderates the relationship between PKA and ADRS in OHCs. This result highlights that not all patients are equally ready to embrace technological changes in OHC interactions. The extent to which patients can learn from ADRS varies. Previous research by (Gao et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) has examined the moderating role of customer ability/readiness in AI services and its impact on the value cocreation process. Our study extends this by investigating the moderation of PAR in OHCs, finding that patients with higher PAR benefit more from learning, whereas those with lower PAR struggle to process complex information.\u003c/p\u003e \u003cp\u003eMoreover, PAR moderates the mediated relationship, which contradicts earlier studies by (Nguyen and Negash, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and (Rebelo et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), who controlled for the effect of customer ability/readiness in various contexts of customer participation and citizenship behavior in online communities. However, our study shows that a patient\u0026rsquo;s desire to engage in participation behavior in OHCs significantly depends on their PAR level. For example, cancer patients with higher PAR who interact with search bots better understand the recommendations and are motivated to assist others in their recovery. Thus, our findings offer novel insights for future research on OHCs. Finally, the robustness test of the model via the MLP confirms the relevance of the integrated components, where sensitivity analysis reveals that the PAR and ADRS are the most significant predictors of VCB. Therefore, the importance of these variables should not be overlooked when aiming to achieve improved VCB outcomes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e7.2 Theoretical and Practical Implications\u003c/h2\u003e \u003cp\u003eRecent research in service studies has increasingly emphasized the importance of value cocreation, particularly with respect to the involvement of multiple actors in service environments (Wang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This study offers several significant contributions to the theoretical landscape. First, it addresses the call to explore value cocreation at the micro level(Laukka et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Peng et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), advancing the literature by developing a sustainable environment within online health communities (OHCs). By integrating social assistance theories and value cocreation, this study proposes a moderated mediation model that lays the foundation for empirical investigations, providing substantial insights into optimizing patients' value cocreation behavior (VCB) within OHCs.\u003c/p\u003e \u003cp\u003eSecond, this research extends the application of social assistance theory to the context of human\u0026ndash;AI interactions, considering the growing use of AI-driven tools such as AI-driven recommendation systems (ADRSs) in various sectors, including healthcare and services(Lopez-Barreiro et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Varidel et al., \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The integration of AI-driven systems such as ADRS in OHCs remains an underexplored area. Our findings shed light on the different types of social assistance that patients receive from ADRS and the perceived benefits and challenges posed by these systems, which significantly influence patients' VCB. Unlike earlier studies(Liu et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Tian and Wu, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), which focused primarily on human-based social assistance, our study is the first to empirically examine how ARDS-generated support influences patient participation and contribution in OHCs, thus opening new avenues for research on how AI tools can enhance VCB in digital health contexts.\u003c/p\u003e \u003cp\u003eThird, our study contributes to the growing literature by identifying the pivotal role of patient knowledge acquisition (PKA) in value cocreation. Peer learning has long been recognized as a key factor in value cocreation (Liu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and our study contrasts with earlier research that emphasized belongingness and ethical perceptions in facilitating value cocreation within OHCs. However, the role of PKA, particularly its empirical validation in OHCs, remains underexplored. Our findings show that PKA, facilitated through social assistance from ADRS, significantly enhances patients' VCB. Few studies have explored how PKA, as a form of SAP from AI-driven systems, mediates VCB. We show that PKA empowers patients to share information and assist others, thereby promoting value cocreation within these communities. With these results, we address the ongoing call for research into how AI search bots and recommenders can stimulate user participation in value cocreation within digital environments (Gao et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Latif \u003cem\u003eet al.\u003c/em\u003e, 2025).\u003c/p\u003e \u003cp\u003eFourth, this study advances our understanding of the moderating role of patient ability/readiness (PAR) in influencing PKA and VCB when interacting with the ADRS. Our findings indicate that PAR significantly enhances PKA and engagement in OHCs when patients interact with ADRSs, contributing valuable insights to the OHC literature. The proposed model suggests that while the SAP from the ADRS positively influences PKA, the strength of this relationship is contingent upon varying levels of PAR. Understanding these levels can guide the development of user-centered interventions within OHCs, fostering effective learning and collaboration for value cocreation. Finally, by combining the PLS‒SEM and MLP methods, we provide new insights that highlight the relevance of independent constructs contributing to VCB. Our results show that the MLP is a more effective predictive model, as demonstrated by the low RMSE values from both the training and testing datasets (Leong et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003e7.2.1. Practical Implications\u003c/h2\u003e \u003cp\u003eThis study offers several practical insights for healthcare managers within online health communities (OHCs). First, the findings emphasize that ADRSs have significant potential for improving value cocreation behavior (VCB). Therefore, OHCs should go beyond using search bots solely for technical support and incorporate AI-powered ADRS that provide both emotional and informational support to community members. For example, informational support (IS) from the ADRS helps patients make informed decisions about their health by providing relevant, timely information. Additionally, ADRS can facilitate community discussions on health-related topics, allowing patients to receive diverse feedback from other members. This study further indicates that when an ADRS offers emotional support (ES), it creates empathetic responses that validate patients' feelings, encouraging them to share personal experiences and emotions within the community. While SAP from ADRS alleviates the burden on human moderators, OHCs should strike a balance between human interaction and automated support to address members\u0026rsquo; concerns effectively.\u003c/p\u003e \u003cp\u003eSecond, the study revealed that patient knowledge acquisition (PKA) mediates the relationship between ADRS and VCB. This suggests that OHC managers should focus on patient education when AI-driven systems are integrated. By offering personalized resources and information tailored to patients' health needs, OHCs can enhance PKA. For example, personalized guidance from the ADRS, which responds to specific health-related queries, boosts patients' confidence in their ability to participate in discussions and advocate for others within the community.\u003c/p\u003e \u003cp\u003eThird, our findings underscore that SAP from ADRSs plays a critical role in fostering PKA and promoting value cocreation within OHCs. However, the effectiveness of ADRS may vary depending on patient ability/readiness (PAR). OHC managers should account for different levels of PAR when designing AI-based support systems. By assessing PAR levels and adopting personalized approaches, search bot responses can be tailored to patients' comprehension levels. The complexity of the language used by ADRS affects how well patients understand and engage with the information. Healthcare managers should ensure that ADRS simplify complex health-related content. Furthermore, OHCs should create safe spaces where patients with lower PAR feel comfortable seeking clarification about their health issues. Patients with higher PAR are more likely to engage with ADRS and actively participate in value cocreation activities within the community.\u003c/p\u003e \u003cp\u003eFinally, this study illustrates how the indirect effect of the SAP from the ADRS on VCB through the PKA is influenced by the PAR. Specifically, the SAP from the ADRS significantly impacts VCB through the PKA when the PAR is high. However, this indirect effect becomes nonsignificant when PAR is low. The identification of this boundary condition provides a deeper understanding of the relationship between ADRS and patients' abilities, contributing to the social assistance literature. These findings also extend current research by highlighting the conditional indirect pathway through which SAPs from the ADRS, PKA, and PAR collectively influence VCB.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"8. Conclusion and Limitations","content":"\u003cp\u003eThis study demonstrates that AI-driven recommendation systems (ADRS) significantly enhance value cocreation behavior (VCB) within online health communities (OHCs) by fostering patient learning and promoting trust-based engagement. The findings reveal that trust in ADRS is the most critical determinant of both learning and cocreation, with perceived risk and privacy concerns acting as barriers to participation. Furthermore, health anxiety positively moderates the relationship between trust and learning, and patient readiness (PAR) strengthens the indirect effect of trust on VCB through learning. By integrating PLS-SEM and MLP, we provide a robust analytical framework that captures both causal and nonlinear relationships, ensuring the reliability and generalizability of the model.\u003c/p\u003e \u003cp\u003eFrom a theoretical perspective, this research extends the understanding of AI-enabled value cocreation by integrating insights from trust theory, learning mechanisms, and the privacy calculus framework within healthcare contexts. Practically, the findings suggest that AI systems designed with transparency, privacy assurance, and emotional sensitivity can significantly enhance patient engagement and collaborative behaviors. Healthcare organizations and ADRS developers should focus on improving user trust and readiness while ensuring that AI tools are empathetic, explainable, and privacy-conscious, empowering patients to actively participate in their digital health journeys.\u003c/p\u003e \u003cp\u003eHowever, several limitations should be considered. First, the study relied on reported data from questionnaire-based surveys, and future research could benefit from incorporating objective data directly from patient interactions within OHCs to improve accuracy. Second, while this study focused on OHCs in China, the findings may not be generalizable to other regions. Future studies should examine the model in diverse geographical contexts to assess its global applicability. Additionally, this study focused primarily on younger patients, so exploring age-related differences in VCB would offer valuable insights. Finally, while PLS-SEM and MLP were effective in this study, future research could explore other machine learning techniques to increase the predictive accuracy and gain a more nuanced understanding of the factors influencing VCB within OHCs. These limitations open several avenues for future research, and addressing them would further contribute to advancing the understanding of ADRS in healthcare and its role in fostering VCB.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eConflicts of Interest:\u003c/h2\u003e \u003cp\u003eThe authors declare no conflict of interest.\u003c/p\u003e \u003ch2\u003eFunding:\u003c/h2\u003e \u003cp\u003eThe authors received no financial support for the research and/or authorship of this article.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e \u003cp\u003eThe data used in this study are available from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAggarwal, C.C. (2015), \u0026ldquo;Outlier Analysis\u0026rdquo;, \u003cem\u003eData Mining\u003c/em\u003e, Springer International Publishing, pp. 237\u0026ndash;263, doi: 10.1007/978-3-319-14142-8_8.\u003c/li\u003e\n\u003cli\u003eAhuja, H. and Shree, D. (2025), \u0026ldquo;Patient-to-patient value co-creation in online health communities: a systematic literature review and research agenda\u0026rdquo;, \u003cem\u003eInternational Journal of Pharmaceutical and Healthcare Marketing\u003c/em\u003e, Emerald Publishing, Vol. 19 No. 1, pp. 107\u0026ndash;138, doi: 10.1108/IJPHM-05-2024-0046.\u003c/li\u003e\n\u003cli\u003eAin, F.U., Siddiqui, A., Mustafa, A., Memon, F. and Kausar, F. (2025), \u0026ldquo;AI-DRIVEN DECISION SUPPORT SYSTEMS IN PUBLIC HEALTH ADMINISTRATION\u0026rdquo;, \u003cem\u003eJournal of Medical \u0026amp; Health Sciences Review\u003c/em\u003e, Insightful Education Research Institute, Vol. 2 No. 3, doi: 10.62019/WAPQJ864.\u003c/li\u003e\n\u003cli\u003eBagheri, M., Bagheritabar, M., Alizadeh, S., Parizi, M. (Sam) S., Matoufinia, P. and Luo, Y. (2025), \u0026ldquo;Machine-Learning-Powered Information Systems: A Systematic Literature Review for Developing Multi-Objective Healthcare Management\u0026rdquo;, \u003cem\u003eApplied Sciences (Switzerland)\u003c/em\u003e, Multidisciplinary Digital Publishing Institute (MDPI), 1 January, doi: 10.3390/app15010296.\u003c/li\u003e\n\u003cli\u003eBahit, M.C., Korjian, S., Chi, G., Daaboul, Y., Castro Molina, A., Jiang, G., Dangas, C., \u003cem\u003eet al.\u003c/em\u003e (2025), \u0026ldquo;Understanding the Burden of Myocardial Infarction and Patient Preferences for Treatment: A Real-World Study Assessing Patients\u0026rsquo; Perspectives via an Online Survey\u0026rdquo;, \u003cem\u003eJournal of Patient-Centered Research and Reviews\u003c/em\u003e, Aurora Health Care, Inc., Vol. 12 No. 4, pp. 234\u0026ndash;244, doi: 10.17294/2330-0698.2146.\u003c/li\u003e\n\u003cli\u003eChen, L., Baird, A. and Straub, D. (2019), \u0026ldquo;Fostering participant health knowledge and attitudes: an econometric study of a chronic disease-focused online health community\u0026rdquo;, \u003cem\u003eJournal of Management Information Systems\u003c/em\u003e, Taylor \u0026amp; Francis, Vol. 36 No. 1, pp. 194\u0026ndash;229.\u003c/li\u003e\n\u003cli\u003eDralus, G., Mazur, D., Kusznier, J. and Dralus, J. (2023), \u0026ldquo;Application of artificial intelligence algorithms in multilayer perceptron and Elman networks to predict photovoltaic power plant generation\u0026rdquo;, \u003cem\u003eEnergies\u003c/em\u003e, MDPI, Vol. 16 No. 18, p. 6697.\u003c/li\u003e\n\u003cli\u003eFebri Ramanda, M. Ari Prayogo, Bagus Dwi Saputra and Muhammad Labib Jundillah. (2025), \u0026ldquo;Exploring User Experience in Adopting AI-Based Information Systems in Healthcare Environments\u0026rdquo;, \u003cem\u003eJurnal Ilmiah Sistem Informasi Dan Ilmu Komputer\u003c/em\u003e, Politeknik Pratama Purwokerto, Vol. 5 No. 2, pp. 517\u0026ndash;529, doi: 10.55606/JUISIK.V5I2.1437.\u003c/li\u003e\n\u003cli\u003eGao, B., Wang, J., Ding, X. and Guo, Y. (2024), \u0026ldquo;The Pitfalls of Review Solicitation: Evidence from a Natural Experiment on TripAdvisor\u0026rdquo;, \u003cem\u003eHttps://Doi.Org/10.1287/Mnsc.2023.01006\u003c/em\u003e, INFORMS, Vol. 71 No. 2, pp. 1671\u0026ndash;1691, doi: 10.1287/MNSC.2023.01006.\u003c/li\u003e\n\u003cli\u003eGao, L., Li, G., Tsai, F., Gao, C., Zhu, M. and Qu, X. (2023), \u0026ldquo;The impact of artificial intelligence stimuli on customer engagement and value co-creation: the moderating role of customer ability readiness\u0026rdquo;, \u003cem\u003eJournal of Research in Interactive Marketing\u003c/em\u003e, Emerald Publishing, Vol. 17 No. 2, pp. 317\u0026ndash;333, doi: 10.1108/JRIM-10-2021-0260.\u003c/li\u003e\n\u003cli\u003eGlikson, E. and Woolley, A.W. (2020), \u0026ldquo;Human Trust in Artificial Intelligence: Review of Empirical Research\u0026rdquo;, \u003cem\u003eHttps://Doi.Org/10.5465/Annals.2018.0057\u003c/em\u003e, Academy of Management AnnalsBriarcliff Manor, NY, Vol. 14 No. 2, pp. 627\u0026ndash;660, doi: 10.5465/ANNALS.2018.0057.\u003c/li\u003e\n\u003cli\u003eGrassini, E., Buzzi, M., Leporini, B. and Vozna, A. (2025), \u0026ldquo;A systematic review of chatbots in inclusive healthcare: insights from the last 5 years\u0026rdquo;, \u003cem\u003eUniversal Access in the Information Society\u003c/em\u003e, Springer, Vol. 24 No. 1, pp. 195\u0026ndash;203.\u003c/li\u003e\n\u003cli\u003eGuan, R., Rakovic, M., Chen, G. and Ga\u0026scaron;evic, D. (2025), \u0026ldquo;How educational chatbots support self-regulated learning? A systematic review of the literature\u0026rdquo;, \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, Springer, Vol. 30 No. 4, pp. 4493\u0026ndash;4518.\u003c/li\u003e\n\u003cli\u003eHair, J. and Alamer, A. (2022), \u0026ldquo;Partial Least Squares Structural Equation Modeling (PLS-SEM) in second language and education research: Guidelines using an applied example\u0026rdquo;, \u003cem\u003eResearch Methods in Applied Linguistics\u003c/em\u003e, Elsevier, Vol. 1 No. 3, p. 100027.\u003c/li\u003e\n\u003cli\u003eHair, J.F., Risher, J.J., Sarstedt, M. and Ringle, C.M. (2019), \u0026ldquo;When to use and how to report the results of PLS-SEM\u0026rdquo;, \u003cem\u003eEuropean Business Review\u003c/em\u003e, Emerald Publishing Limited, Vol. 31 No. 1, pp. 2\u0026ndash;24.\u003c/li\u003e\n\u003cli\u003eHair, J.F., Singh, S., Kaur, R. and Dana, L.P. (2022), \u0026ldquo;A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM)\u0026rdquo;, \u003cem\u003eSAGE Publication\u003c/em\u003e, SAGE Publication, Inc., pp. 1\u0026ndash;352.\u003c/li\u003e\n\u003cli\u003eHarkness, J.A., Braun, M., Edwards, B., Johnson, T.P., Lyberg, L.E., Mohler, P.P., Pennell, B.-E., \u003cem\u003eet al.\u003c/em\u003e (2010), \u003cem\u003eSurvey Methods in Multinational, Multiregional, and Multicultural Contexts\u003c/em\u003e, John Wiley \u0026amp; Sons.\u003c/li\u003e\n\u003cli\u003eHenseler, J., Ringle, C.M. and Sarstedt, M. (2015), \u0026ldquo;A new criterion for assessing discriminant validity in variance-based structural equation modeling\u0026rdquo;, \u003cem\u003eJournal of the Academy of Marketing Science\u003c/em\u003e, Springer, Vol. 43 No. 1, pp. 115\u0026ndash;135.\u003c/li\u003e\n\u003cli\u003eJannach, D. and Adomavicius, G. (2016), \u0026ldquo;Recommendations with a purpose\u0026rdquo;, \u003cem\u003eRecSys 2016 - Proceedings of the 10th ACM Conference on Recommender Systems\u003c/em\u003e, Association for Computing Machinery, Inc, pp. 7\u0026ndash;10, doi: 10.1145/2959100.2959186;JOURNAL:JOURNAL:ACMCONFERENCES;PAGEGROUP:STRING:PUBLICATION.\u003c/li\u003e\n\u003cli\u003eKadir, I., Abdallaoui, A. and Elazhari, K. (2025), \u0026ldquo;Architectural optimization of a multilayer perceptron (MLP) neural network enhanced by the Levenberg\u003cins cite=\"mailto:Editor\" datetime=\"2025-12-09T16:42\"\u003e-\u003c/ins\u003e\u003cdel cite=\"mailto:Editor\" datetime=\"2025-12-09T16:42\"\u003e-\u003c/del\u003eMarquardt algorithm for predicting relative humidity: application to Tangier, Morocco\u0026rdquo;, \u003cem\u003eWater SA\u003c/em\u003e, Water Research Commission (WRC), Vol. 51 No. 3, pp. 279\u0026ndash;287.\u003c/li\u003e\n\u003cli\u003eKetchen, D.J. (2013), \u0026ldquo;A Primer on Partial Least Squares Structural Equation Modeling\u0026rdquo;, \u003cem\u003eLong Range Planning\u003c/em\u003e, Elsevier BV, Vol. 46 No. 1\u0026ndash;2, pp. 184\u0026ndash;185, doi: 10.1016/j.lrp.2013.01.002.\u003c/li\u003e\n\u003cli\u003eKock, N. (2015), \u0026ldquo;Common method bias in PLS-SEM: A full collinearity assessment approach\u0026rdquo;, \u003cem\u003eInternational Journal of E-Collaboration (Ijec)\u003c/em\u003e, IGI Global, Vol. 11 No. 4, pp. 1\u0026ndash;10.\u003c/li\u003e\n\u003cli\u003eKock, N. (2020), \u0026ldquo;Harman\u0026rsquo;s single factor test in PLS-SEM: Checking for common method bias\u0026rdquo;, \u003cem\u003eData Analysis Perspectives Journal\u003c/em\u003e, Vol. 2 No. 2, pp. 1\u0026ndash;6.\u003c/li\u003e\n\u003cli\u003eAl Kuwaiti, A., Nazer, K., Al-Reedy, A., Al-Shehri, S., Al-Muhanna, A., Subbarayalu, A.V., Al Muhanna, D., \u003cem\u003eet al.\u003c/em\u003e (2023), \u0026ldquo;A review of the role of artificial intelligence in healthcare\u0026rdquo;, \u003cem\u003eJournal of Personalized Medicine\u003c/em\u003e, MDPI, Vol. 13 No. 6, p. 951.\u003c/li\u003e\n\u003cli\u003eLatan, H., Hair Jr, J.F., Noonan, R. and Sabol, M. (2023), \u0026ldquo;Introduction to the partial least squares path modeling: Basic concepts and recent methodological enhancements\u0026rdquo;, \u003cem\u003ePartial Least Squares Path Modeling: Basic Concepts, Methodological Issues and Applications\u003c/em\u003e, Springer, pp. 3\u0026ndash;21.\u003c/li\u003e\n\u003cli\u003eLatif, M.S. and Wang, J.-J. (2025), \u0026ldquo;The moderating role of face on value co-creation behavior and co-creation attitude in online health communities\u0026rdquo;, \u003cem\u003eAslib Journal of Information Management\u003c/em\u003e, Emerald Publishing Limited, Vol. 77 No. 3, pp. 464\u0026ndash;487.\u003c/li\u003e\n\u003cli\u003eLatif, M.S., Wang, J.J., Shahzad, M. and Mursil, M. (2024), \u0026ldquo;The impact of social support chatbots on patients\u0026rsquo; value co-creation behavior in online health communities: a moderated mediation model\u0026rdquo;, \u003cem\u003eInternet Research\u003c/em\u003e, Emerald Publishing Limited, Vol. 35 No. 3, pp. 1126\u0026ndash;1153, doi: 10.1108/INTR-08-2023-0666.\u003c/li\u003e\n\u003cli\u003eLaukka, E., Tuunanen, T., Jansson, M., Vanhanen, M., Hirvonen, N., Palukka, J., Vesinurm, M., \u003cem\u003eet al.\u003c/em\u003e (2025), \u0026ldquo;Value Cocreation and Codestruction in Digital Health Services: Protocol for a Systematic Review\u0026rdquo;, \u003cem\u003eJMIR Research Protocols\u003c/em\u003e, JMIR Publications Inc., Toronto, Canada, Vol. 14 No. 1, p. e63015.\u003c/li\u003e\n\u003cli\u003eLeong, L.-Y., Hew, T.-S., Ooi, K.-B., Tan, G.W.-H. and Koohang, A. (2025), \u0026ldquo;An SEM-ANN approach-guidelines in information systems research\u0026rdquo;, \u003cem\u003eJournal of Computer Information Systems\u003c/em\u003e, Taylor \u0026amp; Francis, Vol. 65 No. 6, pp. 706\u0026ndash;737.\u003c/li\u003e\n\u003cli\u003eLeskovec, Jurij., Rajaraman, Anand. and Ullman, J.D.. (2020), \u0026ldquo;Mining of massive datasets\u0026rdquo;, Cambridge University Press, p. 553.\u003c/li\u003e\n\u003cli\u003eLi, L., Pongsakornrungsilp, S., Pongsakornrungsilp, P. and Jiang, Y. (2024), \u0026ldquo;The Influence of Social Support on Customer Engagement in Live Streaming Commerce: The Mediating Role of Perceived Value.\u0026rdquo;, \u003cem\u003eTEM Journal\u003c/em\u003e, Vol. 13 No. 4.\u003c/li\u003e\n\u003cli\u003eLi, M. and Tuunanen, T. (2022), \u0026ldquo;Information Technology\u0026ndash;Supported value Co-Creation and Co-Destruction via social interaction and resource integration in service systems\u0026rdquo;, \u003cem\u003eThe Journal of Strategic Information Systems\u003c/em\u003e, Elsevier, Vol. 31 No. 2, p. 101719.\u003c/li\u003e\n\u003cli\u003eLiu, M., Zou, X., Chen, J. and Ma, S. (2022), \u0026ldquo;Comparative analysis of social support in online health communities using a word co-occurrence network analysis approach\u0026rdquo;, \u003cem\u003eEntropy\u003c/em\u003e, MDPI, Vol. 24 No. 2, p. 174.\u003c/li\u003e\n\u003cli\u003eLiu, S., Xiao, W., Fang, C., Zhang, X. and Lin, J. (2020), \u0026ldquo;Social support, belongingness, and value co-creation behaviors in online health communities\u0026rdquo;, \u003cem\u003eTelematics and Informatics\u003c/em\u003e, Elsevier, Vol. 50, p. 101398.\u003c/li\u003e\n\u003cli\u003eLopez-Barreiro, J., Garcia-Soidan, J.L., Alvarez-Sabucedo, L. and Santos-Gago, J.M. (2024), \u0026ldquo;Artificial intelligence-powered recommender systems for promoting healthy habits and active aging: A systematic review\u0026rdquo;, \u003cem\u003eApplied Sciences\u003c/em\u003e, MDPI, Vol. 14 No. 22, p. 10220.\u003c/li\u003e\n\u003cli\u003eMiner, A.S., Laranjo, L. and Kocaballi, A.B. (2020), \u0026ldquo;Chatbots in the fight against the COVID-19 pandemic\u0026rdquo;, \u003cem\u003eNPJ Digital Medicine\u003c/em\u003e, Nature Publishing Group UK London, Vol. 3 No. 1, p. 65.\u003c/li\u003e\n\u003cli\u003eMiyazaki, A.D. and Haderlie, T.C. (2025a), \u0026ldquo;Emotional AI as a facilitator of co-created service value in a health-care ecosystem: a collective framework\u0026rdquo;, \u003cem\u003eJournal of Services Marketing\u003c/em\u003e, Emerald Publishing Limited, doi: 10.1108/JSM-01-2025-0041.\u003c/li\u003e\n\u003cli\u003eNaga, J., Quidato, M., Pascual, Ma.C.G.P. and Ebardo, R. (2025), \u0026ldquo;Engagement Dynamics in Online Mental Health Communities: The Role of Trust and Social Support in Young Adults\u0026rsquo; Attitudes and Norms\u0026rdquo;, \u003cem\u003eSAGE Open\u003c/em\u003e, SAGE Publishing, Vol. 15 No. 3, doi: 10.1177/21582440251376147.\u003c/li\u003e\n\u003cli\u003eNguyen, T.T.V. and Negash, Y.T. (2024), \u0026ldquo;The Impact of Value Co-Creation Readiness on Customer Experience and Revisit Intention in the Hospitality Industry.\u0026rdquo;, \u003cem\u003eInternational Journal of Information \u0026amp; Management Sciences\u003c/em\u003e, Vol. 35 No. 3.\u003c/li\u003e\n\u003cli\u003eNohutlu, Z.D., Englis, B.G., Groen, A.J. and Constantinides, E. (2023), \u0026ldquo;Innovating with the customer: Co-creation motives in online communities\u0026rdquo;, \u003cem\u003eInternational Journal of Electronic Commerce\u003c/em\u003e, Taylor \u0026amp; Francis, Vol. 27 No. 4, pp. 523\u0026ndash;557.\u003c/li\u003e\n\u003cli\u003eOlawade, D.B., Wada, O.Z., Odetayo, A., David-Olawade, A.C., Asaolu, F. and Eberhardt, J. (2024), \u0026ldquo;Enhancing mental health with Artificial Intelligence: Current trends and future prospects\u0026rdquo;, \u003cem\u003eJournal of Medicine, Surgery, and Public Health\u003c/em\u003e, Elsevier, Vol. 3, p. 100099, doi: 10.1016/J.GLMEDI.2024.100099.\u003c/li\u003e\n\u003cli\u003eOsei-Frimpong, K., Wilson, A. and Lemke, F. (2018), \u0026ldquo;Patient co-creation activities in healthcare service delivery at the micro level: The influence of online access to healthcare information\u0026rdquo;, \u003cem\u003eTechnological Forecasting and Social Change\u003c/em\u003e, Elsevier, Vol. 126, pp. 14\u0026ndash;27.\u003c/li\u003e\n\u003cli\u003e\u0026ldquo;(PDF) Collaborative Filtering Recommender Systems\u0026rdquo;. (n.d.)., available at: https://www.researchgate.net/publication/200121027_Collaborative_Filtering_Recommender_Systems (accessed 9 December 2025).\u003c/li\u003e\n\u003cli\u003ePeng, Y., Wu, T., Chen, Z. and Deng, Z. (2022), \u0026ldquo;Value cocreation in health care: systematic review\u0026rdquo;, \u003cem\u003eJournal of Medical Internet Research\u003c/em\u003e, JMIR Publications Toronto, Canada, Vol. 24 No. 3, p. e33061.\u003c/li\u003e\n\u003cli\u003ePodsakoff, P.M., Podsakoff, N.P., Williams, L.J., Huang, C. and Yang, J. (2024), \u0026ldquo;Common method bias: It\u0026rsquo;s bad, it\u0026rsquo;s complex, it\u0026rsquo;s widespread, and it\u0026rsquo;s not easy to fix\u0026rdquo;, \u003cem\u003eAnnual Review of Organizational Psychology and Organizational Behavior\u003c/em\u003e, Annual Reviews, Vol. 11 No. 1, pp. 17\u0026ndash;61.\u003c/li\u003e\n\u003cli\u003eQiu, X., Sun, T., Xu, Y., Shao, Y., Dai, N. and Huang, X. (2021), \u0026ldquo;Pre-trained Models for Natural Language Processing: A Survey\u0026rdquo;, \u003cem\u003eScience China Technological Sciences\u003c/em\u003e, Springer Verlag, Vol. 63 No. 10, pp. 1872\u0026ndash;1897, doi: 10.1007/s11431-020-1647-3.\u003c/li\u003e\n\u003cli\u003eRamadan, O.M.E., Elsharkawy, N.B., Hafiz, A.H., Katooa, N.E., Abunar, A., Attallah, D.M.A., Dewan, M., \u003cem\u003eet al.\u003c/em\u003e (2025), \u0026ldquo;Neonatal nurses\u0026rsquo; e-health literacy and technology-mediated clinical practice: a cross-sectional analysis of digital health competencies and practice patterns\u0026rdquo;, \u003cem\u003eBMC Nursing\u003c/em\u003e, BioMed Central, Vol. 24 No. 1, doi: 10.1186/S12912-025-03839-7.\u003c/li\u003e\n\u003cli\u003eRebelo, A., Varela-Neira, C. and Ruzo-Sanmart\u0026iacute;n, E. (2024), \u0026ldquo;Boosting Customers\u0026rsquo; Co-Creation in Open-Source Software Environments: The Role of Innovativeness and a Sense of Community\u0026rdquo;, \u003cem\u003eJournal of Theoretical and Applied Electronic Commerce Research\u003c/em\u003e, MDPI, Vol. 19 No. 3, pp. 2476\u0026ndash;2496.\u003c/li\u003e\n\u003cli\u003eRicci, F., Rokach, L. and Shapira, B. (2022), \u0026ldquo;Recommender Systems Handbook: Third Edition\u0026rdquo;, \u003cem\u003eRecommender Systems Handbook: Third Edition\u003c/em\u003e, Springer US, pp. 1\u0026ndash;1060, doi: 10.1007/978-1-0716-2197-4/COVER.\u003c/li\u003e\n\u003cli\u003eRingle, C.M., Wende, S. and Becker, J. (2024), \u0026ldquo;SmartPLS 4. B\u0026ouml;nningstedt: SmartPLS. http s\u0026rdquo;.\u003c/li\u003e\n\u003cli\u003eRoppelt, J.S., Jenkins, A., Kanbach, D.K., Kraus, S. and Jones, P. (2025), \u0026ldquo;Effective adoption of artificial intelligence in healthcare: A multiple case study\u0026rdquo;, \u003cem\u003eJournal of Decision Systems\u003c/em\u003e, Taylor \u0026amp; Francis, Vol. 34 No. 1, p. 2458883.\u003c/li\u003e\n\u003cli\u003eSanaji, S., Handriana, T. and Usman, I. (2022), \u0026ldquo;Antecedents and consequences of value co-creation in online shopping\u0026rdquo;, \u003cem\u003eBISMA (Bisnis Dan Manajemen)\u003c/em\u003e, Universitas Negeri Surabaya, Vol. 15 No. 1, pp. 20\u0026ndash;39, doi: 10.26740/BISMA.V15N1.P20-39.\u003c/li\u003e\n\u003cli\u003eSchroer, H.W. and Just, C.L. (2023), \u0026ldquo;Feature engineering and supervised machine learning to forecast biogas production during municipal anaerobic co-digestion\u0026rdquo;, \u003cem\u003eACS Es\u0026amp;t Engineering\u003c/em\u003e, ACS Publications, Vol. 4 No. 3, pp. 660\u0026ndash;672.\u003c/li\u003e\n\u003cli\u003eSharma, A., Lin, I.W., Miner, A.S., Atkins, D.C. and Althoff, T. (2023), \u0026ldquo;Human\u0026ndash;AI collaboration enables more empathic conversations in text-based peer-to-peer mental health support\u0026rdquo;, \u003cem\u003eNature Machine Intelligence\u003c/em\u003e, Nature Publishing Group UK London, Vol. 5 No. 1, pp. 46\u0026ndash;57.\u003c/li\u003e\n\u003cli\u003eSwan, E.L., Peltier, J.W. and Dahl, A.J. (2024), \u0026ldquo;Artificial intelligence in healthcare: the value co-creation process and influence of other digital health transformations\u0026rdquo;, \u003cem\u003eJournal of Research in Interactive Marketing\u003c/em\u003e, Emerald Publishing Limited, Vol. 18 No. 1, pp. 109\u0026ndash;126.\u003c/li\u003e\n\u003cli\u003eSzeto, M.D., Hook Sobotka, M., Woolhiser, E., Parmar, P., Wu, J., Alhanshali, L. and Dellavalle, R.P. (2024), \u0026ldquo;PatientsLikeMe and online patient support communities in dermatology\u0026rdquo;, \u003cem\u003eJMIR Dermatology\u003c/em\u003e, JMIR Publications Toronto, Canada, Vol. 7, p. e50453.\u003c/li\u003e\n\u003cli\u003eTan, Q., Tan, J. and Gao, X. (2024), \u0026ldquo;How does the online innovation community climate affect the user\u0026rsquo;s value co-creation behavior: The mediating role of motivation\u0026rdquo;, \u003cem\u003ePLOS ONE\u003c/em\u003e, Public Library of Science, Vol. 19 No. 4, p. e0301299, doi: 10.1371/JOURNAL.PONE.0301299.\u003c/li\u003e\n\u003cli\u003eTao, Guannan; Yan, Sanjiu, ; and Zhou, Y. (2025), \u0026ldquo;Analysis of the Driving Factors for Fan economy Growth on Social Media Platforms Based on the Push-Pull Theory: Taking the Largest Platform in China as an Example Standard-Nutzungsbedingungen\u0026rdquo;.\u003c/li\u003e\n\u003cli\u003eThe Annenberg Public Policy Center of the University of Pennsylvania. (2025), \u0026ldquo;Many in U.S. Consider AI-Generated Health Information Useful and Reliable | The Annenberg Public Policy Center of the University of Pennsylvania\u0026rdquo;, July, available at: https://www.annenbergpublicpolicycenter.org/many-in-u-s-consider-ai-generated-health-information-useful-and-reliable/ (accessed 2 December 2025).\u003c/li\u003e\n\u003cli\u003eTian, X.-F. and Wu, R.-Z. (2022), \u0026ldquo;Determining factors affecting the users\u0026rsquo; participation of online health communities: an integrated framework of social capital and social support\u0026rdquo;, \u003cem\u003eFrontiers in Psychology\u003c/em\u003e, Frontiers Media SA, Vol. 13, p. 823523.\u003c/li\u003e\n\u003cli\u003eTrigka, M. and Dritsas, E. (2025), \u0026ldquo;A comprehensive survey of deep learning approaches in image processing\u0026rdquo;, \u003cem\u003eSensors\u003c/em\u003e, MDPI, Vol. 25 No. 2, p. 531.\u003c/li\u003e\n\u003cli\u003eVanderhout, S., Smith, M., Lee-Foon, N., Doherty-Kirby, A., Fleming, R., Grant, D., Grenier, A., \u003cem\u003eet al.\u003c/em\u003e (2025), \u0026ldquo;Co-creating an everyday language illustration of learning health systems alongside patient, caregiver, and community partners\u0026rdquo;, \u003cem\u003eLearning Health Systems\u003c/em\u003e, Wiley, doi: 10.1002/LRH2.70033.\u003c/li\u003e\n\u003cli\u003eVaridel, M., An, V., Hickie, I.B., Cripps, S., Marchant, R., Scott, J., Crouse, J.J., \u003cem\u003eet al.\u003c/em\u003e (2025), \u0026ldquo;Causal AI Recommendation System for Digital Mental Health: Bayesian Decision-Theoretic Analysis\u0026rdquo;, \u003cem\u003eJournal of Medical Internet Research\u003c/em\u003e, JMIR Publications Toronto, Canada, Vol. 27, p. e71305.\u003c/li\u003e\n\u003cli\u003eVickery, S.K., Castel, M.J., Narayanan, S. and Nicolae, M.L. (2025), \u0026ldquo;A Model of Hospital Patient Engagement for Value Co-Creation: Does It Affect Performance? A U.S. Hospital Industry Study\u0026rdquo;, \u003cem\u003eJournal of Operations Management\u003c/em\u003e, Wiley, doi: 10.1002/JOOM.70019.\u003c/li\u003e\n\u003cli\u003eVinh, V.M. and Hung, N.G. (2025), \u0026ldquo;AI-Driven Success: Exploring the Role of Technology Readiness and Artificial Intelligence Integration on Digital Business Performance\u0026rdquo;, \u003cem\u003eJournal of Economics, Finance and Management Studies\u003c/em\u003e, Everant Journals, Vol. 08 No. 09, doi: 10.47191/JEFMS/V8-I9-44.\u003c/li\u003e\n\u003cli\u003eWang, C., Zhao, X. and Hong, J. (2024), \u0026ldquo;A Meta-Analysis of the Effects of Interaction on Value Co-Creation in Online Collaborative Innovation Communities Based on the Service Ecosystem Framework\u0026rdquo;, \u003cem\u003eBehavioral Sciences\u003c/em\u003e, MDPI, Vol. 14 No. 12, p. 1177.\u003c/li\u003e\n\u003cli\u003eWang, J., Yao, T. and Wang, Y. (2023), \u0026ldquo;Patient engagement as contributors in online health communities: the mediation of peer involvement and moderation of community status\u0026rdquo;, \u003cem\u003eBehavioral Sciences\u003c/em\u003e, MDPI, Vol. 13 No. 2, p. 152.\u003c/li\u003e\n\u003cli\u003eWei, J. (2025a), \u0026ldquo;Analysis of the Behavioral Performance and Social Support of Patients in Online Health Communities From User Profile Perspectives: Comparative Study\u0026rdquo;, \u003cem\u003eJournal of Medical Internet Research\u003c/em\u003e, JMIR Publications Toronto, Canada, Vol. 27, p. e68074.\u003c/li\u003e\n\u003cli\u003eXiao Fei, Y., Loh, X.M., Seduram, L.A. /P and Tan, G.W.H. (2025), \u0026ldquo;How is brand love shaped through virtual event value co-creation and sponsorship? A multi-method approach\u0026rdquo;, \u003cem\u003eJournal of Research in Interactive Marketing\u003c/em\u003e, Emerald Publishing, doi: 10.1108/JRIM-11-2024-0516/1255978.\u003c/li\u003e\n\u003cli\u003eXiao, L. and Han, X. (2025), \u0026ldquo;Enhancing Patient Value Co-creation via AI-enabled Cases Disclosure: The Role of Uncertainty Reduction, Patient Empowerment and Risk Aversion\u0026rdquo;, \u003cem\u003eSocial Science \u0026amp; Medicine\u003c/em\u003e, Elsevier, p. 118822.\u003c/li\u003e\n\u003cli\u003eYang, B., Lu, W., Xuan, Y., Hao, C. and Huang, X. (2025), \u0026ldquo;The influences of social support expressed from doctors and disclosed from peers on patient decision-making: an analysis from the online health community\u0026rdquo;, \u003cem\u003eDental Science Reports\u003c/em\u003e, Nature Portfolio, Vol. 15 No. 1, doi: 10.1038/S41598-024-85023-6.\u003c/li\u003e\n\u003cli\u003eYang, J., Ren, J., Zhao, M., Chen, H. and Wang, H. (2023), \u0026ldquo;The impact of value co-creation on consumer citizenship behavior: Based on consumer perspective\u0026rdquo;, \u003cem\u003eFrontiers in Psychology\u003c/em\u003e, Frontiers Media S.A., Vol. 13, p. 1110011, doi: 10.3389/FPSYG.2022.1110011/BIBTEX.\u003c/li\u003e\n\u003cli\u003eYi, Y. and Gong, T. (2013), \u0026ldquo;Customer value co-creation behavior: Scale development and validation\u0026rdquo;, \u003cem\u003eJournal of Business Research\u003c/em\u003e, Elsevier, Vol. 66 No. 9, pp. 1279\u0026ndash;1284, doi: 10.1016/J.JBUSRES.2012.02.026.\u003c/li\u003e\n\u003cli\u003eYin, J., Zhao, Y., Ma, Z., Chen, S. and Li, M. (2025), \u0026ldquo;Customer citizenship behavior and customer perceived value in China: the mediating role of value co-creation experience\u0026rdquo;, \u003cem\u003eHumanities and Social Sciences Communications\u003c/em\u003e, Palgrave.\u003c/li\u003e\n\u003cli\u003eZhang, Z. and Lu, M. (2025), \u0026ldquo;When engagement meets departure: how social capital, social engagement and social support impact churn within online health communities\u0026rdquo;, \u003cem\u003eAslib Journal of Information Management\u003c/em\u003e, Emerald Publishing Limited, doi: 10.1108/AJIM-04-2024-0350.\u003c/li\u003e\n\u003cli\u003eZhao, H. (2025), \u0026ldquo;The impact of social support on individual health knowledge creation in online patient communities: the mediating role of attitude\u0026rdquo;, \u003cem\u003eJournal of Knowledge Management\u003c/em\u003e, Emerald Publishing Limited, Vol. 29 No. 3, pp. 937\u0026ndash;967.\u003c/li\u003e\n\u003cli\u003eZhao, L. and Liu, T. (2025), \u0026ldquo;Exploring generational and educational disparities in online health information seeking in China: The moderating role of internet use\u0026rdquo;, \u003cem\u003eDigital Health\u003c/em\u003e, SAGE Publications Inc., Vol. 11, doi: 10.1177/20552076251393378.\u003c/li\u003e\n\u003cli\u003eZhou, P., Zhao, Y., Xiao, S. and Zhao, K. (2022), \u0026ldquo;The impact of online health community engagement on lifestyle changes: A serially mediated model\u0026rdquo;, \u003cem\u003eFrontiers in Public Health\u003c/em\u003e, Frontiers Media SA, Vol. 10, p. 987331.\u003c/li\u003e\n\u003cli\u003eZou, J. and Shao, Y. (2022), \u0026ldquo;A Study on Factors Affecting the Value Co-Creation Behavior of Customers in Sharing Economy: Take Airbnb Malaysia as an Example\u0026rdquo;, \u003cem\u003eSustainability 2022, Vol. 14, Page 12678\u003c/em\u003e, Multidisciplinary Digital Publishing Institute, Vol. 14 No. 19, p. 12678, doi: 10.3390/SU141912678.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Northeastern University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"AI-Driven Recommendation Systems, Online Health Communities, Value Co-Creation Behavior, Patient Knowledge Acquisition, Patient Ability/Readiness, Social Assistance","lastPublishedDoi":"10.21203/rs.3.rs-8413621/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8413621/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study explores how AI-driven recommendation systems (ADRS) influence patients' participation in value cocreation behavior (VCB) within online health communities (OHCs), aiming to explore the mechanisms behind the fostering of sustainable OHCs. Using a moderated mediation model grounded in social assistance theories and value cocreation, survey data from 450 respondents were analysed via partial least squares structural equation modelling (PLS-SEM) and a multilayer perceptron (MLP) neural network model. The findings reveal that social assistance perceived (SAP) from ADRS positively influences VCB, both indirectly and directly through patient knowledge acquisition (PKA), and that the indirect effect is more robust when patient ability/readiness (PAR) is high. The MLP analysis further confirms the robustness of the model and the substantial moderating role of PAR. This research introduces a novel framework that sheds light on the role of the ADRS in enhancing SAP and PKA, ultimately driving VCB in OHCs. The findings provide practical insights for developing user-centric interventions to foster effective learning and collaboration in AI-driven health platforms, ensuring long-term sustainability and engagement in OHCs.\u003c/p\u003e","manuscriptTitle":"The Impact of AI-Driven Recommendation Systems on Patients’ Value Co-Creation Behavior in Online Health Communities: A Moderated Mediation Model","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-30 00:54:37","doi":"10.21203/rs.3.rs-8413621/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a63b640d-8380-427c-b96d-caf0042e1e06","owner":[],"postedDate":"December 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":60005008,"name":"Marketing"}],"tags":[],"updatedAt":"2025-12-30T00:54:37+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-30 00:54:37","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8413621","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8413621","identity":"rs-8413621","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}