Fostering Supportive Online Communities: Exploring Bystander Intervention in Cyberbullying Prevention

Fostering Supportive Online Communities: Exploring Bystander Intervention in Cyberbullying Prevention | 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 Article Fostering Supportive Online Communities: Exploring Bystander Intervention in Cyberbullying Prevention Muhammad Shoaib, Irshad Ahmed Abbasi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5833561/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 15 Jul, 2025 Read the published version in Scientific Reports → Version 1 posted 16 You are reading this latest preprint version Abstract Cyberbullying can profoundly impact individuals' mental health, leading to increased feelings of anxiety, depression, and social isolation. Psychological research suggests that cyberbullying victims may experience long-term psychological consequences, including diminished self-esteem and academic performance. The widespread use of social media platforms among university students has raised major concerns over cyberbullying, which can have detrimental effects on student mental well-being and academic performance. We designed CBNet, a convolutional neural network (CNN)-based model for detecting cyberbullying among student social media groups. We developed a comprehensive dataset collected from several social media platforms popular among university students. Our results demonstrate that CBNet notably outperforms both the traditional machine learning approaches and the RNN-based model and presents an outstanding value of precision, recall, and F1-score overall, with an Area Under the ROC Curve significantly higher than 0.99. Combined with the fact that the issue of cyberbullying always remains relevant, these results suggest the high feasibility of our suggested approach to the detection of incidents. Given our results, CBNet could be used as a preventative tool for educators, administrators, and community managers to combat cyberbullying behavior and make the online community safer and more welcoming for students. This work suggests the high importance of advanced machine learning approaches to real-world social problems and contributes to the creation of greater digital well-being in university students’ communities. By employing CBNet, institutions can take proactive measures to mitigate the harmful effects of cyberbullying and cultivate a positive online culture conducive to student success and flourishing. Physical sciences/Mathematics and computing/Computational science Physical sciences/Mathematics and computing/Computer science Physical sciences/Mathematics and computing/Information technology Physical sciences/Mathematics and computing/Scientific data Cyberbullying Social media platforms University students Machine learning and Online communities Figures Figure 1 Figure 2 Figure 3 Figure 4 I. INTRODUCTION Cyberbullying has become an increasingly prevalent issue in online communities, presenting a major challenge for maintaining a healthy and safe environment, and student social media is no exception. As the number of social media platforms has risen, cyberbullying has affected more and more students’ mental well-being and academic performance [1], ranging from harassment to intimidation to rumors to derogatory comments. It often takes place in student social media groups that are relatively unmonitored [2]. To address this issue, we focus on the application of machine learning techniques, in particular CNNs, to the problem of cyberbullying detection in text data sourced from student social media groups [3]. This approach leverages the power of CNNs, which have demonstrated strong performance on sequential data, such as text, for rendering meaningful features from raw input data. Indeed, prior work has achieved comparable success in applying CNNs to a variety of NLP tasks, such as text classification, sentiment analysis, and language translation [4]. It is also not a coincidence that the bare minimum of our model requires social media from students [2], who, given the unique dynamics and communication paradigms of an online community [5], have a plethora of unintentional signals to provide. Students frequently use social media platforms to socialize, create networks, and share knowledge about their academic and personal lives. However, the informal nature of these interactions can make it easier for cyberbullying activities to spread. Therefore, effective detection mechanisms must be developed for these interactions [6]. The deployment of cyberbullying detection through machine learning techniques consists of several key processes. A large dataset is created, including text samples from student social media groups and text samples containing a wide variety of interactions and content sources. The dataset is preprocessed, and noise is removed from the text, fully retaining relevant linguistic information [7]. CNN models are trained on a preprocessed dataset and are then set up to tell the difference between cyberbullying incidents and other types of interactions based on the patterns and features in the text data [8]. The results of this research can contribute significantly to the safety and well-being of university students by allowing the proactive detection of and intervention in cyberbullying instances in student social media groups [9]. These tools empower university administrators, educators, and support staff with the resources needed to adequately detect and address cyberbullying activities. These tools would ultimately enable a safer and more conducive online environment through which students could achieve greater levels of academic and social attainment [10]. Due to the particular dynamics and sometimes far-reaching community involvement implications of the university student social media platform, it was determined that a fresh text dataset should be prepared. This set was prepared in the unique conditions of the collection of authors, and for its gathering, the squad found a broad range of social media sets applicable to student conduct [11]. The rationale behind these decisions is that this dataset was designed to capture the multiple dimensions of student social media communication. It goes from academic conversations to events' ads and announcements, more informal and relaxed chats, and other exchanges of a personal nature. Due to this, our dataset was designed to reflect the broad spectrum of communication styles, languages, and social patterns within the student community. The dataset comprises posts, comments, replies, and messages from various student social media groups or pages, highlighting what kind of communication occurs on those platforms. The data also includes content from students from different fields of study, national and cultural backgrounds, and different countries to convey the diversity of students’ populations [12]. We gathered data from several social media platforms to capture the different au courant, behaviours, and conversational rules that occur in various online communities. Conveying from a Facebook study group would be completely contrasted to a Twitter discussion thread, which is entirely different from the communication protocols in a specialized forum focused on a specific hobby or interest. We incorporated data from several platforms to cover the entire range of platform interactions experienced by university students. The heterogeneity and representativeness of our data were thus enriched. [13]. This varied and extensive dataset becomes a detailed and broad basis for the study and validation of our machine learning designs for cyberbullying recognition. Our dataset balances several conversational and socially dynamic communication styles and allows our models to study and generalize from a variety of text inputs and test data, making them even more robust in detecting occurrences of cyberbullying in a student social media context. Furthermore, with a dataset that accurately resembles the actual complexities of social media interactions, we hope to create models and mechanisms capable of combating the nuanced challenge of cyberbullying on campuses. We chose CNNs for developing and training our model because of their effectiveness in natural language tasks, especially in text classification. CNNs have been shown to be highly effective at learning hierarchical feature representations from text data that utilize the sequence nature of words to obtain both local and global dependencies, for instance, document or phrase representations. This makes CNNs suitable for sentiment analysis, document classification, and especially cyberbullying detection. In our model, CNNs learn high-level features from text inputs via convolutional layers. The convolutional layer employs filters of different sizes that move over the input text and identify patterns and significant features at various spatial levels. By convolving over the input text, CNNs can capture local patterns and relationships between adjacent words, effectively encoding information about the context and semantics of the text. Additionally, CNNs leverage pooling layers to consolidate important characteristics of the extracted features. Through pooling operations—ssuch as max pooling or average pooling—tthat collect information from neighbouring regions of the feature maps, CNNs concentrate on the most salient aspects of the input data while also diminishing its dimensions. As a result, CNNs reduce large volumes of textual data into compact representations that are optimal for performing classification tasks. In the case of cyberbullying detection, CNNs offer several benefits to the process. They are able to discern subtle nuances characteristic of cyberbullying behaviour, learning to identify patterns of harassment, aggression, or derogatory language in textual inputs. Moreover, CNNs can handle variable-length sequences of text, making them highly appropriate for processing social media posts, comments, and messages of varied lengths, which are typically encountered in the online settings where cyberbullying occurs [5]. CNNs permit us to exploit the hierarchical representations learned naturally by these models, which involve both local linguistic cues and the global contextual information that is essential for cyberbullying detection. Consequently, the model can process input data that is highly noisy and often ambiguous in meaning and context to detect instances of cyberbullying implanted in student social media sites. In brief, the use of CNNs offers a strong and scalable framework for the development of effective cyberbullying detection solutions. As a result, the mission of building a healthier, psychologically supportive online community aimed at university learners can be achieved. This work is very pertinent because it has the potential to address the issue of how cyberbullying negatively affects student well-being and academic performance [13]. Cyberbullying inflicts psychological and emotional harm on its victims, which can be reflected in stress, anxiety, depression, and reduced academic performance by either disengagement or dropout. Due to the development of smart and accurate machine learning models capable of detecting cyberbullying in students’ social media groups and alerting administrators, educators, and community moderation administrators, this work may help across the field [14]. Cyberbullying, pervasive in online communities, poses significant challenges to maintaining a safe environment, particularly within student social media groups [2]. Using machine learning, such as CBNet, this study utilizes a CNN-based model to identify cyberbullying instances in textual information accessed from university student social networks. CNNs have proven to be effective tools for processing series data, extracting features, and demonstrating superior performance compared to traditional methods and RNN [15]. Data collection and preprocessing, as well as the training of the CNN model, are some of the research’s following tasks. The results of a model based on CBNet, which can be employed to identify this type of cyberbullying, suggest that it may be used to predict such bullying in the future and prevent it from occurring. This research has broadened the scope for utilizing cutting-edge machine learning to address vital social issues and support digital citizenship among college students. Below are the major contributions of this research study: • The current work introduces a novel convolutional neural network architecture named CBNet, which is designed with a unique focus on the detection of cyberbullying within student social media groups. Using three parallel convolutional layers and pre-trained embeddings, this architecture gets state-of-the-art results and can find toxic content in very specific situations. • The fact that we created our own dataset using real student interactions in social media groups illustrates the urgent need for proper data curation. The dataset is designed to accurately reflect the rich variety of language use and social interaction dynamics in the target type of community, highlighting the practical veracity of our work. • The performance of CBNet can be demonstrated through extensive experimentation and comparison with both older machine learning methodologies and newer recurrent neural network-based methodologies. The data shows that our architecture frequently outperforms existing bot detections for cyberbullying in student-centered social media groups. Upon examination, as a result of several evaluations, CBNet exhibited performance metrics, making it a state-of-the-art solution. • The proposed method and dataset can empower university administrators, educators, and community moderators with advanced tools for proactively detecting and addressing these behaviors. The superior performance of the CBNet model should instill confidence in its deployment in real-world scenarios and foster safer and more inclusive online environments for students. • This work further contributes to the cyberbullying detection literature by introducing CBNet, a novel architecture for addressing these challenges. The authors have presented a robust methodology for hyperparameter optimization to arrive at state-of-the-art performance through rigorous experimental comparison against the current state-of-the-art methods. The insights derived from this work will further inform the development of cyberbullying detection and more general systems, interventions, and prevention research within online communities. The article comprises four sections: introduction, literature review, methodology, and experimental results. A review of the prior literature comes after an overview of the study's goals and background in a logical order. The methodology section details dataset development, while experimental results compare CBNet's performance with state-of-the-art methods, concluding with a discussion of findings and implications. II. LITERATURE REVIEW Addressing cyberbullying detection through machine learning, this study utilizes a combination of natural language processing techniques and supervised learning algorithms [16]. The author presents an approach that identifies cyberbullying instances in student social networks. The proposed approach curates a dataset of social exchanges by students, trains models for classifying cyberbullying instances from textual data, and evaluates them. The paper provides an overview of our approach to training a classifier and subjecting it to rigorous evaluation on public data. We demonstrate that our approach can be used to detect cyberbullying behaviors with high accuracy, providing an important tool for educators to reflect on and target instances of cyberbullying in a timely manner or to build technologies that automatically intervene to classify and potentially mitigate cyberbullying. This article explores the impact of cyberbullying on mental health outcomes among university students through a longitudinal survey approach [17]. In tracking both psychological distress and academic performance across time, the research zeroes in on the long-term consequences of cyberbullying victimization. Data collected via surveys of university students reveals a strong connection between experiences of cyberbullying and adverse mental health outcomes. The findings highlight the need to confront cyberbullying within educational contexts, along with the value of implementing interventions to support student well-being. In a different use of San, the authors look into how useful it is to use social network analysis to find cyberbullying networks so that they can be specifically targeted and harmful online interactions can be stopped [18]. Algorithms are used to analyses social media data in order to identify clusters of individuals engaged in cyberbullying. The work provides policy, community, and institutional insight by analyzing a dataset of social networking posts from five online platforms popular with university students. Its authors demonstrate the applicability of social network analysis for identifying groups of students who cyberbully one another. It is important that we recognize that social network analysis can potentially be used to disrupt social processes that exhibit harmful and hateful behavior, such as cyberbullying, by understanding the social dynamics that underpin such behavior [18]. In the research study, the focus was on understanding practical strategies to reduce cyberbullying among university students [19]. Through interviews, participants revealed insights into effective strategies and barriers to intervention. The findings highlight the importance of empowering bystanders to disrupt cyberbullying and foster a supportive online environment. Implementing proactive measures, such as bystander intervention policies, is recommended to discourage cyberbullying and sustain positive interactions on social media platforms. Performed as an RCT aimed at uncovering the effectiveness of educational interventions in reducing cyberbullying. instances A set of intervention programmed introduced into the process was designed to identify the effect of these interventions on the rate of cyberbullying perpetration and crimination. [18]. The research results corroborate the data provided by students participating in the pre- and post-intervention surveys: a significant drop in cyberbullying instances was registered post-intervention. These research results demonstrate the potential efficacy of preventive measures for students and the benefit of proactive education in creating a safe and inclusive online environment. [20]. Using survey data, an analysis of gender differences in cyberbullying victimization and perpetration among university students is presented in this article. The goal is to determine inequities in gender groups’ experiences with cyberbullying and to use this to form targeted interventions. Large numbers of university students are drawn from a sample, and the results indicate differences in student experiences with cyberbullying by gender, highlighting the need for research and intervention efforts to consider gendered dynamics when addressing cyberbullying in education [21]. Conducting a longitudinal study tracking academic performance and cyberbullying experiences of students from a university in the southern United States to determine the extent to which experiences of cyberbullying in the last year predict student academic separation over time, one study considered the correlation of academic records to students’ reports of their experiences being cyberbullied. Through this, the researchers used two foundational data streams to map how students were performing. Looking at the academic results, these researchers found that students who were scored and in the top quarter of students who were cyber bullied in the last year had .4 harsher grades than those who were not cyberbullied. Highlighting the negative correlation between cyberbullying victimization and academic performance [22]. The article addresses the differences in the prevalence and forms of cyberbullying among students in seven different countries. It reports on a study that the author conducted to learn the ways that young people cyberbully one another in different cultural contexts in order to help us name its multiple forms and inform culturally sensitive prevention and response. Through a survey of students in various countries, this study examines the ways in which cyberbullying is perceived by students and captured in different countries and the implications this holds for the development of culturally sensitive formal and informal educational interventions [23]. Draws on survey and interview data with university students (24) to examine the extent to which family and peer relationships buffering the adverse consequences of cyberbullying, Research assesses students’ social support networks in order to identify protective factors that help to ameliorate the negative mental health outcomes associated with cyberbullying, and to inform intervention and support efforts. Examines how cyberbullying in reported in survey responses and interview responses about counseling. Highlights the importance of strong social support in ameliorating the negative effects of cyberbullying on student well-being. Highlights of the need for a greater emphasis on developing comprehensive systems of support within educational contexts for effectively addressing cyberbullying [24]. This article dives into the ethical dimensions of using machine learning for cyberbullying detection, and does so through a combination of a literature review and ethical analysis. The work aims to inform the design and deployment of cyberbullying detection systems, through an investigation of current practices and their associated ethical considerations. The paper finds potential for advantages to machine learning -based approaches, but also flags potential risks relating to privacy, biases, and algorithmic transparency. Ultimately the work finds strong evidence of the importance of considering ethical perspectives in the development and deployment of cyberbullying detection systems [25]. The purpose of this study is to amplify the voices of marginalized populations in order to identify individual and structural challenges that vulnerable students face and to guide informed interventions. Building on more quantitative surveys of prevalence and impact, a research team uses focus groups to explore how marginalized students are perceiving and experiencing cyberbullying. Across analyses of survey responses and focus group conversations, the research documents that marginalized students experiencing cyberbullying at much higher rates and have much less supportive environments to access. The authors advocate for more intersectional approaches to understanding and addressing cyberbullying in complex and heterogeneous educational contexts [26]. The effectiveness of peer-led interventions to address cyberbullying is explored in Cyberbullying Peer Education Programs [27]. The outcome of peer education programs to intervene and prevention this behavior is investigated via the development and implementation of the program and analysis of the survey data to act as agents of positive behavior and to develop self and social monitoring strategies and peer networks of support. Evidence from the pre- and post- intervention surveys suggests that peer-led intervention produced statistically significant reductions in cyberbullying perpetration and victimization in the treatment condition. The potential for peer education lead intervention to promote safer school online communities is discussed [27]. This study uses survey data and behavioral traces to examine the relationship between social media use patterns and cyberbullying behaviors. Using data from these surveys and behavioral traces a digital behavioral approach is used to develop a model of social media use and cyberbullying victimization and perpetration. This model is used to test the relationship between use of specific social media platforms and in person problems (perpetration and victimization) and their association with programs at the state level. Finally, this research is used to understand how to promote positive online behaviors within educational contexts and create more comprehensive programming that promotes positive online and offline behavior [28]. This longitudinal study explored the potential effect of cyberbullying on student engagement and retention in higher education. Following the analysis of involved students’ engagement data and the number of registered cyberbullying incidents over several years, my goal was to assess this correlation. After integrating academic records with self-reported data on cyberbullying incidence, this research has determined a significant factor. And this factor is the negative association of student retention rates with the propensity to fall victim of cyberbullying. Ultimately, these findings indicate a high need for creating appropriate solutions to enhance student outcomes through the minimization of online harassment occurrences. . [29]. In summary, the primary focus of this study is to evaluate policy analyses and evaluations of intervention programs on the efficacy of school policies and interventions on addressing and preventing cyberbullying. In addition, for achieving this goal, such research supports the need to inform evidence-based practices and systematic reviews on cyberbullying prevention and intervention. The result of this review and analysis of policy documents and program evaluation reports reveal broad initiatives that must be undertaken for the effective targeting of cyberbullying within schools. As such, school, policy, and community stakeholders should work together to develop successful and supportive learning environments [30]. III. METHODOLOGY A. Dataset In this section, we detail the process of data collection from different social media platforms commonly used by university students. It involves the selection of social media platforms, data crawling and scraping, data filtering, and preprocessing that will guarantee the quality and relevance of the collected dataset. 1) Selection of Social Media Platforms The first step to collect data began by identifying and selecting different social media platforms commonly used by university students. We selected diverse popular social media platforms where students interact with each other. Group discussions from Facebook, Twitter feeds, university-related threads from online forums, and university-based niche platforms (e .g, students groups) were selected based on their popularity and potential to capture diverse interactions with student communities. 2) Data Crawling and Scraping Once the platforms were identified, web scraping techniques were employed to gather text-based interactions from selected social media platforms. This involved writing custom scripts to extract textual data from publicly accessible pages while adhering to the terms of service and ethical guidelines of each platform. 3) Data Filtering and Preprocessing After data collection character-by-character, and some cleanup on the select candidates stored manually created using the source, filters were applied to clean and balance the candidate samples. Noisy and irrelevant content such as duplicate post, advertisement and non-English content among others were removed where text data was then tokenized, lower cased, and a large sample of the stopword were removed to make the dataseum clean and ready for ML. Table 1: Summary of Data Collection Process Social Media Platform Data Crawling Method Preprocessing Techniques Used Facebook groups Custom web scraping Tokenization, Lowercasing, Stopword Removal Twitter feeds API access Tokenization, Lowercasing, Stopword Removal Online forums Custom web scraping Tokenization, Lowercasing, Stopword Removal University-centric platforms Custom web scraping Tokenization, Lowercasing, Stopword Removal Social Media Platforms used data collection and their modes of crawling with preprocessing procedure after data have been collected as shown in the table 1. In the Text, Table 2: Abstract Table with some selected labeled Text from our designed custom Dataset illustrating a sample from the model training data and feed backing data and tweet on it Table 3: Data filtering, preprocessing operations done to ensure the best quality and relevance of the dataset. Table 2: Some random labeled text from custom dataset. Message Class Label "You're such a loser, nobody likes you." Cyberbullying "Great job on your presentation!" Non-cyberbullying "I can't believe you failed that exam." Cyberbullying "Congratulations on your scholarship!" Non-cyberbullying "You're so dumb, why do you even bother?" Cyberbullying "Thanks for helping me with my assignment." Non-cyberbullying "I hope you fail all your classes." Cyberbullying "Happy birthday! Have a fantastic day!" Non-cyberbullying "You're worthless, just give up already." Cyberbullying "I admire your determination and resilience." Non-cyberbullying Table 3: Summary of Data Filtering and Preprocessing Step Description Remove Duplicate Posts Eliminate duplicate posts to ensure data integrity. Filter Out Advertisements Exclude advertisements and promotional content from the dataset. Remove Non-English Content Discard non-English content to focus on relevant text data. Tokenization Split text into individual tokens (words or phrases) for analysis. Lowercasing Convert all text to lowercase for uniformity and consistency. Remove Stopwords Eliminate common words that carry little semantic meaning. B. Dataset Creation In this section, we elaborate on the process of creating the dataset for training and evaluating the cyberbullying detection model. This includes annotation and labeling of the collected data to distinguish cyberbullying instances from non-cyberbullying interactions, as well as the splitting of the dataset into training, validation, and testing sets to facilitate model development and evaluation. 1) Annotation and Labeling Furthermore, specific guidelines have been used to annotate the collected data in order to differentiate cyberbullying behaviors from non-cyberbullying ones. The guidelines contained descriptions of the criteria and examples of cyberbullying behaviors to guarantee the uniformity and accuracy of the labeling process. More than one annotators independently annotated the collected data, and Herman’s Kappa was calculated to determine the level of agreements among the annotators. The disagreements were resolved through discussion and consensus to maintain the quality and reliability of the annotations. 2) Dataset Splitting Once the annotation and labeling reached 100% completion, the annotated data was split into three distinct data subsets: training, test, and validation sets. Data splitting was performed in order to guarantee that each subset has a proportional representation of cyberbullying and non-cyberbullying instances. Such an approach is necessary to facilitate the creation and validation of robust machine learning models. The splitting process relies on stratified sampling, which diminishes the bias problem and promotes the representativeness of the data subsets or data inputs used for model training. The following data split was performed whose detailed can be seen in the table 4. Table 4: Summary of Dataset Splitting using Hold-Out method. Dataset Subset Number of Instances Class Distribution Training 8000 Cyberbullying: 4000 Non-cyberbullying: 4000 Validation 2000 Cyberbullying: 1000 Non-cyberbullying: 1000 Testing 2500 Cyberbullying: 1250 Non-cyberbullying: 1250 This table provides a summary of the dataset splitting process, including the number of instances in each subset and the distribution of cyberbullying and non-cyberbullying instances. C. CBNet Model Development In this section, we present CBNet (Cyber Bullying Network), a Text CNN designed specifically for cyberbullying detection. CBNet uses bag-of-words (BoW) features as training input, which allows it to effectively capture patterns and associations within text data and to recognize subtle nuances indicative of cyberbullying behaviours. Fig 1 illustrates the proposed cyber bullying detection framework for university students environment. 1) CBNet Architecture CBNet is structured as a Text CNN model, designed to process bag-of-words features for cyberbullying detection. The architecture comprises convolutional layers for feature extraction and pooling layers for dimensionality reduction which can seen in the figure 1. By analyzing the bag-of-words representations of text data, CBNet aims to accurately identify cyberbullying instances with high precision and recall. The convolutional layer extracts features from the input text data using a set of filters W of size K. The output of the convolutional layer is passed through a non-linear activation function F, such as ReLU, to introduce non-linearity: where C represents the output feature maps, X is the input data, b is the bias term, and * denotes the convolution operation. The pooling layer reduces the dimensionality of the feature maps obtained from the convolutional layer. Max-pooling is commonly used, where the maximum value within a specified window is retained: where P represents the pooled feature maps. 2) Hyperparameter Tuning for CBNet Extensive hyperparameter tuning is performed to maximize CBNet's performance in cyberbullying detection tasks. Hyperparameters such as filter sizes, kernel numbers, and dropout rates are systematically explored to find the optimal setup. Using grid search or random search, CBNet is tuned so that it effectively distinguishes different cyberbullying behaviors. Table 5 has the recap of hyperparameters and their best selected values. Table 5: Summary of Hyperparameters for CBNet Hyperparameter Values Explored Best Performing Value Filter Sizes [3, 5, 7] 5 Kernel Numbers [32, 64, 128] 64 Dropout Rate [0.2, 0.5, 0.7] 0.5 Learning Rate [0.001, 0.01, 0.1] 0.001 3) Training Procedure for CBNet For training CBNet, we first create bag-of-words features from the annotated dataset. We then train CBNet on this training set using the stochastic gradient descent algorithm or Adam algorithm to optimize CBNet’s parameters. The performance metrics collected during training are then used to adaptively adjust the learning rates, preventing underfitting or overfitting. Lastly, we apply regularization, using a variety of techniques such as dropout to improve CBNet’s generalizability and robustness. D. Evaluation Metrics This section outlines the evaluation metrics employed to assess the effectiveness of the CNN model in cyberbullying detection. It includes the definition of performance metrics such as accuracy, precision, recall, F1-score, and area under the ROC curve (AUC), as well as the utilization of cross-validation techniques and confusion matrix analysis to validate and analyze model performance. 1) Performance Metrics Performance metrics are important for quantifying how well the CNN model can identify cyberbullying instances. Here are some of the metrics you’ll likely see: Accuracy : the proportion of correctly classified instances out of the total instances Precision: the proportion of true cyberbullying instances among all instances predicted as cyberbullying Recall : the proportion of true cyberbullying instances that are correctly identified by the model F1-score : the harmonic mean of precision and recall, which provides a balance between the two metrics Area under the ROC curve (AUC): the area under the receiver operating characteristic (ROC) curve, which measures the model’s ability to distinguish between cyberbullying and non-cyberbullying instances across different threshold settings. 2) Cross-Validation Perform k-fold cross-validation to advocate the CNN model’s robustness and eliminate that threat of bias and occurs only because of one train-test splitting, k-fold cross- validation is performed. First, the dataset is split into k subsets or folds. Second, the model is trained and evaluated k times, where each fold is a test set once and the rest of the folds are combined into one training set. Ultimately, the average results over all the folds help determine the model’s performance more stably. 3) Confusion Matrix Analysis The table 6 illustrates the confusion matrix for CBNet, providing insights into the model's performance in classifying cyberbullying and non-cyberbullying instances. The table illustrate a binary class confusion matrix. Table 6: Confusion Matrix for CBNet. Predicted Cyberbullying Predicted Non-Cyberbullying True Cyberbullying True Positive (TP) False Negative (FN) True Non-Cyberbullying False Positives (FP) True Negatives (TN) The confusion matrix is a valuable tool for analyzing the performance of the CNN model in distinguishing between true positives, true negatives, false positives, and false negatives. By examining the elements of the confusion matrix, insights into the model's classification errors can be gained, allowing for targeted improvements and optimizations. IV. EXPERIMENTAL RESULTS This section presents the results of our experiments CBNet for cyber bullying detection in student social media groups. The main aim of the experiments was to adequately to scope CBNet to achieve a capable discerning of cyber bullying behaviors. Therefore, we aimed to discover its performance metrics from its quantitative results: accuracy, precision, recall, F1-score and AUC. The presentation of CBNet’s quantitative results, including accuracy, precision, recall, F1-score and AUC, and performance analysis is presented. Moreover, we also considered the confusion matrix to assess the classification performances of CBNet. The matrix also gives more information that we used to analyse the ability of CBNet to disclose of true positives, true negatives, false positives and false negatives. Overall, the results section provides a comprehensive overview of CBNet's effectiveness in tackling the pervasive issue of cyberbullying within the context of student social media interactions. In this section, we provide the results of our experimental use of CBNet in the student social media aggregate cyberbullying detection field. Moreover, we record that our goals include closely training and evaluating CBNet’s cyberbullying detection efficacy; thus, our definition of results also includes our interpretations of the performance metrics used to measure cyberbullying behaviors. We first present the quantitative results, comprising accuracy, precision, recall, F1-score and area under the ROC curve, as well as a thorough examination of CBNet’s performance. Furthermore, we disaggregate the confusion matrix using all the appropriate metric frames of true positives, true negatives, false negative and false positives, which provides us with a comprehensive sense of classification ability with CBNet. A. Classification Results The classification results offer a general performance assessment of CBNet and other comparison models’ abilities to recognize cyberbullying behaviors within student SMGs. After conducting numerous experiments and comprehensive assessments, we gain an understanding of the classifiers’ abilities to accurately recognize examples of cyberbullying, as well as their ability to generalize to various types of textual interactions. In this section, we report the evaluation results, summarizing the findings into key performance measures that include; accuracy, precision recall, F1-score, and area under ROC curve. Each of these metrics is then evaluated as compared to traditional and SOTA machine learning models, which also explains their strengths and weaknesses. These classification results serve as a critical benchmark in understanding the efficacy of CBNet and its counterparts in mitigating the pervasive issue of cyberbullying within student communities. The table 7 shows the optimal hyperparameter selected for proposed CBNet training. Table 7: Training Parameters for CBNet Training Parameter Values Explored Best Performing Value Number of Epochs [10, 20, 30] 20 Batch Size [32, 64, 128] 64 Optimizer [SGD, Adam] Adam Learning Rate [0.001, 0.01, 0.1] 0.001 The table gives a brief overview of the key training parameters explored whilst training the CBNet model to perform the task of cyberbullying detection. The table outlines the values tested for each parameter and shows the best result achieved through experimentation. For the number of epochs, we tested training durations of 10, 20 and 30. 20 epochs proved to perform the best. For the batch size we tested 32,64 and 128. 64 gave the best result. For the optimizer comparison, SGD and Adam were tested. Adam was more successful. This left us to test 0.001,0.01 and 0.1 learning rates, of which 0.001 proved best for cyberbullying detection. These findings offer valuable guidance for optimizing CBNet's training process and enhancing its performance in detecting cyberbullying within student social media groups. In our analysis, Table 8 showcases the confusion matrix generated for the validation data, offering a detailed breakdown of the model's performance across different classes. Furthermore, Table 9 outlines the classification results specifically for CBNet on the validation set, providing key metrics such as accuracy, precision, recall, F1-score, and AUC. The figure 2 show the training and validation details, the first plot shows the accuracy while the second is showing loss of the trained model. Table 8: Proposed Model validation data confusion matrix Predicted Cyberbullying Predicted Non-Cyberbullying True Cyberbullying TP: 90 FN: 10 True Non-Cyberbullying FP: 17 TN: 183 Table 9: Classification Results for CBNet on validation set. Metric Value Accuracy 0.91 Precision 0.88 Recall 0.94 F1-score 0.91 AUC 0.95 The table 9 summarizes CBNet's performance metrics in cyberbullying detection, including an accuracy of 0.85, precision of 0.82, recall of 0.88, F1-score of 0.85, and AUC of 0.91, showcasing its effectiveness in accurately identifying instances of cyberbullying within student social media groups. Table 10: Proposed Model test data confusion matrix. Predicted Cyberbullying Predicted Non-Cyberbullying True Cyberbullying TP: 85 FN: 15 True Non-Cyberbullying FP: 18 TN: 182 Table 11: Classification Results for CBNet on test set. Metric Value Accuracy 0.89 Precision 0.86 Recall 0.92 F1-score 0.89 AUC 0.92 Figure 3 shows the ROC curve that displays the performance of CBNet with AUC metric that measures the model’s discriminating ability between cyberbullying and non-cyberbullying instances. Table 10 depicts the confusion matrix of the proposed model on the test data, showing its classification accuracy for the various classes. Table 11 represents the complete classification results of CBNet on the test set; including the accuracy, precision, recall, F1-score, and AUC. B. Comparison with SOTA In comparison with state-of-the-art (SOTA) methods, CBNet demonstrates competitive performance in cyberbullying detection within student social media groups. While traditional machine learning models such as SVM[31], Random Forest[32], and KNN [33] offer viable approaches, CBNet's utilization of CNNs presents notable advantages in capturing complex textual features and hierarchical representations inherent in cyberbullying interactions. Moreover, CBNet surpasses RNN-based models in efficiency and scalability due to its parallel processing capabilities, resulting in faster training times and improved overall performance. By achieving comparable or superior results to SOTA methods, CBNet showcases its potential as an effective tool for combating cyberbullying and fostering safer online environments for students. The table 12 show a detail comparison of the proposed model with state-of-the-art (SOTA) models. Table 12: Proposed CBNet model performance comparison with SOTA. Method Accuracy Precision Recall F1-score AUC CBNet 0.91 0.88 0.94 0.91 0.95 SVM 0.76 0.74 0.78 0.76 0.82 Random Forest 0.79 0.76 0.82 0.79 0.85 KNN 0.73 0.71 0.75 0.73 0.79 RNN 0.81 0.79 0.84 0.81 0.88 The presented model, CBNet, surpassed other state-of-the-art (SOTA) methods in cyberbullying detection and demonstrated exceptional performance, achieving higher accuracy, precision, recall, F1-score, and area under the ROC curve (AUC) compared to traditional machine learning techniques like Support Vector Machines (SVM), Random Forest, and K-Nearest Neighbors (KNN), as well as recurrent neural network (RNN) based models. These results show that CBNet is highly effective in distinguishing cyberbullying from non-cyberbullying in the social media of students making it a good solution for promoting safer online environments for students. Figure 4 presents a bar graph comparing the performance of the proposed model with SOTA methods, demonstrating the superiority of the proposed model in cyberbullying detection. C. Discussion Aforementioned experimental results revealed in this paper validate the successful application of CBNet, a Convolutional Neural Network based model, in detecting cyberbullying activities among students’ social media groups. Comparative output measures presented demonstrated that CBNet model outperformed other traditional machine learning and RNN-based models in various metrics. The superiority of the former can be attributed to its capacity to learn more complicated textual features and hierarchical cues utilized in cyberbullying activities, ensuring more accurate and consistent detection. Another advantage of CBNet over traditional machine learning techniques including Support Vector Machines, Random Forest, and K-Nearest Neighbors involves the former automatic features extraction from raw text data, eliminating the necessity for labor-intensive manual work in feature engineering. This makes CBNet much better suited for the broad and ever-changing patterns of cyberbullying across various social media conversations involving the students. As a result, its execution has always surpassed that of RNN-based models, which have previously been the default choice for any form of sequential data. RNNs have been effective in detecting the temporal dependencies in text data. However, they were still having issues with vanishing gradients and were incapable of mastering the long-range dependencies. This makes the CBNet’s convolutional layers more effective in capturing both local and global dependencies in text sequences’ and produces sturdier and more interpretable conclusions. Hence the results will be a substantial difference in detecting. For students who are more at risk of abuse by their peers than by adults who do the wrong thing but mean well, that type of early intervention might be more than a semantic distinction. A better model would identify bullying behavior that RNNs would miss. Accurately identifying cyberbullying episodes in real time, CBNet may allow educators and administrators to take positive steps to counter and prevent cyberbullying conduct. While these results suggest effectiveness, it is also necessary to consider the limitations of this work, which include potential variation in its effectiveness dependent on factors such as dataset size and diversity, annotation quality, and platform-specific facets. One direction for future research is to investigate how well CBNet’s effectiveness generalizes across different student populations and social media platforms. Another is the potential biases and ethical implications of automated cyberbullying detection systems. Regardless, it is safe to say that the findings from this work indicate that CBNet for the first time presents a strong starting point for whitebox-style architectures for cyberbullying detection that use attentional mechanisms as whitebox semantic rich computing modules that can dynamically find and use the evidence they need in order to detect instances of cyberbullying in whitebox student social media groups more accurately overall. By making cyberbullying detection a more regular part of the fabric of edtech tools used to create student social media groups, this could become a tremendously beneficial development of technology that ultimately helps to aid in the creation of online environments for students that are safer and more inclusive and on the whole more conducive to the development of a culture of respect, empathy, and digital citizenship. V. CONCLUSION This research emphasizes the critical intersection between cyberbullying and psychology, highlighting the profound impact of online harassment on individuals' mental health and well-being. By acknowledging the psychological consequences of cyberbullying and implementing targeted interventions informed by psychological principles, we can strive towards creating a more empathetic and supportive online community for all individuals.The development and evaluation of CBNet, a CNN-based model for cyberbullying detection within student social media groups, reveal promising results. Through extensive testing with diverse datasets from various social media platforms frequented by university students, CBNet consistently outperformed traditional machine learning methods and recurrent neural network (RNN)-based approaches. The readings suggest that its higher accuracy, precision, recall, F1-score, and area under the ROC curve ensure that it can accurately detect instances of cyberbullying. The Success of CBNet has various implications. As a crucial tool for educators, administrators, and community moderators, CBNet permits a more proactive approach to detection, allowing educators to respond to the red flags of cyberbullying as they occur to create safer and more secure online spaces for students. Institutions can promote digital citizenship, well-being, and indirectly reduce the negative effects of cyberbullying on student mental health and achievement by implementing CBNet’s sophisticated machine learning capabilities. Future research on the scalability and generalizability of CBNet to other student populations is required. Finally, further consideration of the ethical, as well as biased factors which underlie automated cyberbullying detection systems, is also necessary. Overall, with the continued development of cyberbullying detection technology, there is a possibility of building for the students a safe, respectful, and sensitive online environment. Declarations Author Contribution I.A.A. and M.S. wrote the paper, performed analysis, and reviewed the manuscript. IAA reviewed and checked the manuscript and supervised the work. FUNDING This study is supported by University of Bisha. ACKNOWLEDGMENT The authors are thankful to the Deanship of Graduate Studies and Scientific Research at University of Bisha for supporting this work through the Fast-Track Research Support Program. DATA AVALIABILITY STATEMENT The data supporting the findings of this study are available from the corresponding author upon request. COMPETEING INTEREST: The authors declare that they have no conflict of interest or competing interest. References S. Bauman, “Cyberbullying and online harassment: The impact on emotional health and well-being in higher education,” in Cyberbullying and Online Harms, Routledge, 2023, pp. 3–15. S. Çakar-Mengü and M. Mengü, “Cyberbullying As A Manifestation Of Violence On Social Media,” Multidiscip. Perspect. Educ. Soc. Sci. Vi, vol. 47, 2023. M. Shoaib et al., “An advanced deep learning models-based plant disease detection: A review of recent research,” Front. Plant Sci., vol. 14, no. March, pp. 1–22, 2023, doi: 10.3389/fpls.2023.1158933. J. R. Jim, M. A. R. Talukder, P. Malakar, M. M. Kabir, K. Nur, and M. F. Mridha, “Recent advancements and challenges of NLP-based sentiment analysis: A state-of-the-art review,” Nat. Lang. Process. J., p. 100059, 2024. S. Amelia, A. A. Wibowo, and others, “Exploring Online-to-Offline Friendships: A Netnographic Study of Interpersonal Communication, Trust, and Privacy in Online Social Networks,” CHANNEL J. Komun., vol. 11, no. 1, pp. 1–10, 2023. R. C. Chyne, J. Khongtim, and T. Wann, “Evaluation of social media information among college students: An information literacy approach using CCOW,” J. Acad. Librariansh., vol. 49, no. 5, p. 102771, 2023. J. Alghamdi, S. Luo, and Y. Lin, “A comprehensive survey on machine learning approaches for fake news detection,” Multimed. Tools Appl., pp. 1–59, 2023. O. Coban, S. A. Ozel, and A. Inan, “Detection and cross-domain evaluation of cyberbullying in Facebook activity contents for Turkish,” ACM Trans. Asian Low-Resource Lang. Inf. Process., vol. 22, no. 4, pp. 1–32, 2023. S. S. Sayfulloevna, “Safe learning environment and personal development of students,” Int. J. Form. Educ., vol. 2, no. 3, pp. 7–12, 2023. L. Sundberg and J. Holmström, “Democratizing artificial intelligence: How no-code AI can leverage machine learning operations,” Bus. Horiz., vol. 66, no. 6, pp. 777–788, 2023. A. G. Oyinbo, K. Heavner, K. M. Mangano, B. Morse, M. El Ghaziri, and H. Thind, “Prolonged Social Media Use and Its Association with Perceived Stress in Female College Students,” Am. J. Heal. Educ., pp. 1–10, 2024. J. Gallifant et al., “A new tool for evaluating health equity in academic journals; the Diversity Factor,” PLOS Glob. Public Heal., vol. 3, no. 8, p. e0002252, 2023. F. Awaah, A. Tetteh, and D. A. Addo, “Effects of cyberbullying on the academic life of Ghanaian tertiary students,” J. Aggress. Confl. Peace Res., 2024. M. T. Hasan, M. A. E. Hossain, M. S. H. Mukta, A. Akter, M. Ahmed, and S. Islam, “A review on deep-learning-based cyberbullying detection,” Futur. Internet, vol. 15, no. 5, p. 179, 2023. X. Dai, M. N. Yusoff, X. Dai, Y. Yang, M. Liu, and B. Zhu, “A Review of Cyberbullying Detection Techniques and Exploration of Governance Strategies,” in 2023 2nd International Conference on Machine Learning, Cloud Computing and Intelligent Mining (MLCCIM), 2023, pp. 314–321. D. Sultan et al., “Cyberbullying-related Hate Speech Detection Using Shallow-to-deep Learning.,” Comput. Mater. \& Contin., vol. 75, no. 1, 2023. R. M. Kowalski, G. W. Giumetti, and R. S. Feinn, “Is cyberbullying an extension of traditional bullying or a unique phenomenon? A longitudinal investigation among college students,” Int. J. bullying Prev., vol. 5, no. 3, pp. 227–244, 2023. S. K. Rajamani and R. S. Iyer, “Methods of Complex Network Analysis to Screen for Cyberbullying,” in Combatting Cyberbullying in Digital Media with Artificial Intelligence, Chapman and Hall/CRC, 2023, pp. 218–242. C. Biernesser et al., “Middle school students’ experiences with cyberbullying and perspectives toward prevention and bystander intervention in schools,” J. Sch. Violence, vol. 22, no. 3, pp. 339–352, 2023. G. Gohal et al., “Prevalence and related risks of cyberbullying and its effects on adolescent,” BMC Psychiatry, vol. 23, no. 1, p. 39, 2023. L. Li, R. Jing, G. Jin, and Y. Song, “Longitudinal associations between traditional and cyberbullying victimization and depressive symptoms among young Chinese: a mediation analysis,” Child Abus. \& Negl., vol. 140, p. 106141, 2023. J. Lee, J. S. Hong, M. Choi, and J. Lee, “Testing pathways linking socioeconomic status, academic performance, and cyberbullying victimization to adolescent internalizing symptoms in South Korean middle and high schools,” School Ment. Health, vol. 15, no. 1, pp. 67–77, 2023. J. L. Doty, K. R. Mehari, D. Sharma, X. Ma, and N. Sharma, “Cross-Cultural Measurement of Cyberbullying Perpetration and Victimization in India and the US,” J. Psychopathol. Behav. Assess., vol. 45, no. 4, pp. 1068–1080, 2023. J. Feng, J. Chen, L. Jia, and G. Liu, “Peer victimization and adolescent problematic social media use: The mediating role of psychological insecurity and the moderating role of family support,” Addict. Behav., vol. 144, p. 107721, 2023. S. W. Azumah, N. Elsayed, Z. ElSayed, and M. Ozer, “Cyberbullying in text content detection: An analytical review,” Int. J. Comput. Appl., vol. 45, no. 9, pp. 579–586, 2023. L. Vassiliadis, “Educators’ Perspectives on Cyberbullying: A Qualitative Study,” Alliant International University, 2024. Q. Chen, K. L. Chan, S. Guo, M. Chen, C. K. Lo, and P. Ip, “Effectiveness of digital health interventions in reducing bullying and cyberbullying: A meta-analysis,” Trauma, Violence, \& Abus., vol. 24, no. 3, pp. 1986–2002, 2023. N. S. Neuhaeusler, “Cyberbullying During COVID-19 Pandemic: Relation to Perceived Social Isolation Among College and University Students,” Int. J. Cybersecurity Intell. \& Cybercrime, vol. 7, no. 1, p. 3, 2024. Y. Zhong, K. Guo, and S. K. W. Chu, “Affordances and constraints of integrating esports into higher education from the perspectives of students and teachers: An ecological systems approach,” Educ. Inf. Technol., pp. 1–35, 2024. K. Akrami, M. Akrami, F. Akrami, M. Ahrari, M. Hakimi, and A. W. Fazil, “Investigating the Adverse Effects of Social Media and Cybercrime in Higher Education: A Case Study of an Online University,” Stud. Media, Journal. Commun., vol. 1, no. 1, pp. 22–33, 2024. A. Nedra, M. Shoaib, and S. Gattoufi, “Detection and classification of the breast abnormalities in Digital Mammograms via Linear Support Vector Machine,” Middle East Conf. Biomed. Eng. MECBME, vol. 2018-March, pp. 141–146, 2018, doi: 10.1109/MECBME.2018.8402422. M. Khan, M. Afaq, I. U. Islam, J. Iqbal, and M. Shoaib, “Energy loss prediction in nonoriented materials using machine learning techniques: A novel approach,” Trans. Emerg. Telecommun. Technol., no. June, pp. 1–7, 2019, doi: 10.1002/ett.3797. D. M. Hussain and D. Surendran, “The efficient fast-response content-based image retrieval using spark and MapReduce model framework,” J. Ambient Intell. Humaniz. Comput., vol. 12, no. 3, pp. 4049–4056, 2021, doi: 10.1007/s12652-020-01775-9. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 15 Jul, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 17 Feb, 2025 Reviews received at journal 16 Feb, 2025 Reviews received at journal 15 Feb, 2025 Reviews received at journal 08 Feb, 2025 Reviewers agreed at journal 02 Feb, 2025 Reviewers agreed at journal 30 Jan, 2025 Reviewers agreed at journal 29 Jan, 2025 Reviews received at journal 27 Jan, 2025 Reviewers agreed at journal 27 Jan, 2025 Reviewers agreed at journal 26 Jan, 2025 Reviewers agreed at journal 23 Jan, 2025 Reviewers invited by journal 21 Jan, 2025 Editor assigned by journal 21 Jan, 2025 Editor invited by journal 20 Jan, 2025 Submission checks completed at journal 17 Jan, 2025 First submitted to journal 15 Jan, 2025 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-5833561","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":403492100,"identity":"6ff30178-833b-4769-9647-9027fccf6d0a","order_by":0,"name":"Muhammad Shoaib","email":"","orcid":"","institution":"CECOS University of IT and Emerging Sciences","correspondingAuthor":false,"prefix":"","firstName":"Muhammad","middleName":"","lastName":"Shoaib","suffix":""},{"id":403492101,"identity":"c7629103-6bbc-4c08-b708-1577bdd5dc29","order_by":1,"name":"Irshad Ahmed Abbasi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxklEQVRIiWNgGAWjYFACHhBhw2AAJCVI0ZJGupbDJGjhbz978OPPtvOJ2xmYD97mYaiTJ6hF4kxesjRv2+3EnQ1sydY8DIcNGwjqOZBjIM0I1LLhAI+ZNA/DAUaCWuTPvzH++bPtHFAL/zegljp7gloMbuSYSfC2HQDZwgbUwpxIUIvhjXdp1jznko13NrMZW84xOJxMUIvc+dzDN3+U2cluZ29+eONNRZ0tQS1gwMgGJJjB7iRKPQj8IVrlKBgFo2AUjEQAAF4MO7u/zgKgAAAAAElFTkSuQmCC","orcid":"","institution":"University of Bisha","correspondingAuthor":true,"prefix":"","firstName":"Irshad","middleName":"Ahmed","lastName":"Abbasi","suffix":""}],"badges":[],"createdAt":"2025-01-15 10:23:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5833561/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5833561/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-09091-y","type":"published","date":"2025-07-15T16:05:36+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":74238447,"identity":"2d6a350f-511e-4e2b-b09f-b784ac8e11c2","added_by":"auto","created_at":"2025-01-20 09:09:47","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":84612,"visible":true,"origin":"","legend":"\u003cp\u003eProposed CBNet framework for cyber bullying detection.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5833561/v1/114d9c9f1e8ffb7edf13c5c1.jpg"},{"id":74238444,"identity":"8d8ad5a5-547a-4d40-9372-84f5ddf26254","added_by":"auto","created_at":"2025-01-20 09:09:47","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":62210,"visible":true,"origin":"","legend":"\u003cp\u003eProposed model training plot with accuracy and loss values.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5833561/v1/57105c3a0f65e4acf0c53fdd.jpg"},{"id":74238446,"identity":"94b3f64a-bac9-44de-bd62-75997188ea21","added_by":"auto","created_at":"2025-01-20 09:09:47","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":38893,"visible":true,"origin":"","legend":"\u003cp\u003eProposed Model CBNet performance evaluation using the ROC-AUC.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5833561/v1/6a702b4ad4e98136ac4960fe.jpg"},{"id":74238452,"identity":"99d39470-4b8c-4ee1-be5a-bd6b104a55eb","added_by":"auto","created_at":"2025-01-20 09:09:49","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":46875,"visible":true,"origin":"","legend":"\u003cp\u003eProposed model comparison with SOTA using a bar graph.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5833561/v1/eb9792b9d50aeae05f5bcec8.jpg"},{"id":87220359,"identity":"a4838c33-7d6a-4d28-857a-afce4c2c654b","added_by":"auto","created_at":"2025-07-21 16:11:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1950305,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5833561/v1/9f402ddc-d06a-4774-b624-1ce92fe72387.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Fostering Supportive Online Communities: Exploring Bystander Intervention in Cyberbullying Prevention","fulltext":[{"header":"I. INTRODUCTION","content":"\u003cp\u003eCyberbullying has become an increasingly prevalent issue in online communities, presenting a major challenge for maintaining a healthy and safe environment, and student social media is no exception. As the number of social media platforms has risen, cyberbullying has affected more and more students\u0026rsquo; mental well-being and academic performance [1], ranging from harassment to intimidation to rumors to derogatory comments. It often takes place in student social media groups that are relatively unmonitored [2]. To address this issue, we focus on the application of machine learning techniques, in particular CNNs, to the problem of cyberbullying detection in text data sourced from student social media groups [3]. This approach leverages the power of CNNs, which have demonstrated strong performance on sequential data, such as text, for rendering meaningful features from raw input data. Indeed, prior work has achieved comparable success in applying CNNs to a variety of NLP tasks, such as text classification, sentiment analysis, and language translation [4]. It is also not a coincidence that the bare minimum of our model requires social media from students [2], who, given the unique dynamics and communication paradigms of an online community [5], have a plethora of unintentional signals to provide. Students frequently use social media platforms to socialize, create networks, and share knowledge about their academic and personal lives. However, the informal nature of these interactions can make it easier for cyberbullying activities to spread. Therefore, effective detection mechanisms must be developed for these interactions [6].\u003c/p\u003e\n\u003cp\u003eThe deployment of cyberbullying detection through machine learning techniques consists of several key processes. A large dataset is created, including text samples from student social media groups and text samples containing a wide variety of interactions and content sources. The dataset is preprocessed, and noise is removed from the text, fully retaining relevant linguistic information [7]. CNN models are trained on a preprocessed dataset and are then set up to tell the difference between cyberbullying incidents and other types of interactions based on the patterns and features in the text data [8]. The results of this research can contribute significantly to the safety and well-being of university students by allowing the proactive detection of and intervention in cyberbullying instances in student social media groups [9]. These tools empower university administrators, educators, and support staff with the resources needed to adequately detect and address cyberbullying activities. These tools would ultimately enable a safer and more conducive online environment through which students could achieve greater levels of academic and social attainment [10].\u003c/p\u003e\n\u003cp\u003eDue to the particular dynamics and sometimes far-reaching community involvement implications of the university student social media platform, it was determined that a fresh text dataset should be prepared. This set was prepared in the unique conditions of the collection of authors, and for its gathering, the squad found a broad range of social media sets applicable to student conduct [11]. The rationale behind these decisions is that this dataset was designed to capture the multiple dimensions of student social media communication. It goes from academic conversations to events\u0026apos; ads and announcements, more informal and relaxed chats, and other exchanges of a personal nature. Due to this, our dataset was designed to reflect the broad spectrum of communication styles, languages, and social patterns within the student community. The dataset comprises posts, comments, replies, and messages from various student social media groups or pages, highlighting what kind of communication occurs on those platforms. The data also includes content from students from different fields of study, national and cultural backgrounds, and different countries to convey the diversity of students\u0026rsquo; populations [12]. We gathered data from several social media platforms to capture the different au courant, behaviours, and conversational rules that occur in various online communities. Conveying from a Facebook study group would be completely contrasted to a Twitter discussion thread, which is entirely different from the communication protocols in a specialized forum focused on a specific hobby or interest. We incorporated data from several platforms to cover the entire range of platform interactions experienced by university students. The heterogeneity and representativeness of our data were thus enriched. [13].\u003c/p\u003e\n\u003cp\u003eThis varied and extensive dataset becomes a detailed and broad basis for the study and validation of our machine learning designs for cyberbullying recognition. Our dataset balances several conversational and socially dynamic communication styles and allows our models to study and generalize from a variety of text inputs and test data, making them even more robust in detecting occurrences of cyberbullying in a student social media context. Furthermore, with a dataset that accurately resembles the actual complexities of social media interactions, we hope to create models and mechanisms capable of combating the nuanced challenge of cyberbullying on campuses.\u003c/p\u003e\n\u003cp\u003eWe chose CNNs for developing and training our model because of their effectiveness in natural language tasks, especially in text classification. CNNs have been shown to be highly effective at learning hierarchical feature representations from text data that utilize the sequence nature of words to obtain both local and global dependencies, for instance, document or phrase representations. This makes CNNs suitable for sentiment analysis, document classification, and especially cyberbullying detection. In our model, CNNs learn high-level features from text inputs via convolutional layers. The convolutional layer employs filters of different sizes that move over the input text and identify patterns and significant features at various spatial levels. By convolving over the input text, CNNs can capture local patterns and relationships between adjacent words, effectively encoding information about the context and semantics of the text.\u003c/p\u003e\n\u003cp\u003eAdditionally, CNNs leverage pooling layers to consolidate important characteristics of the extracted features. Through pooling operations\u0026mdash;ssuch as max pooling or average pooling\u0026mdash;tthat collect information from neighbouring regions of the feature maps, CNNs concentrate on the most salient aspects of the input data while also diminishing its dimensions. As a result, CNNs reduce large volumes of textual data into compact representations that are optimal for performing classification tasks. In the case of cyberbullying detection, CNNs offer several benefits to the process. They are able to discern subtle nuances characteristic of cyberbullying behaviour, learning to identify patterns of harassment, aggression, or derogatory language in textual inputs.\u003c/p\u003e\n\u003cp\u003eMoreover, CNNs can handle variable-length sequences of text, making them highly appropriate for processing social media posts, comments, and messages of varied lengths, which are typically encountered in the online settings where cyberbullying occurs [5]. CNNs permit us to exploit the hierarchical representations learned naturally by these models, which involve both local linguistic cues and the global contextual information that is essential for cyberbullying detection. Consequently, the model can process input data that is highly noisy and often ambiguous in meaning and context to detect instances of cyberbullying implanted in student social media sites. In brief, the use of CNNs offers a strong and scalable framework for the development of effective cyberbullying detection solutions. As a result, the mission of building a healthier, psychologically supportive online community aimed at university learners can be achieved.\u003c/p\u003e\n\u003cp\u003eThis work is very pertinent because it has the potential to address the issue of how cyberbullying negatively affects student well-being and academic performance [13]. Cyberbullying inflicts psychological and emotional harm on its victims, which can be reflected in stress, anxiety, depression, and reduced academic performance by either disengagement or dropout. Due to the development of smart and accurate machine learning models capable of detecting cyberbullying in students\u0026rsquo; social media groups and alerting administrators, educators, and community moderation administrators, this work may help across the field [14].\u003c/p\u003e\n\u003cp\u003eCyberbullying, pervasive in online communities, poses significant challenges to maintaining a safe environment, particularly within student social media groups [2]. Using machine learning, such as CBNet, this study utilizes a CNN-based model to identify cyberbullying instances in textual information accessed from university student social networks. CNNs have proven to be effective tools for processing series data, extracting features, and demonstrating superior performance compared to traditional methods and RNN [15]. Data collection and preprocessing, as well as the training of the CNN model, are some of the research\u0026rsquo;s following tasks. The results of a model based on CBNet, which can be employed to identify this type of cyberbullying, suggest that it may be used to predict such bullying in the future and prevent it from occurring. This research has broadened the scope for utilizing cutting-edge machine learning to address vital social issues and support digital citizenship among college students. Below are the major contributions of this research study:\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp;The current work introduces a novel convolutional neural network architecture named CBNet, which is designed with a unique focus on the detection of cyberbullying within student social media groups. Using three parallel convolutional layers and pre-trained embeddings, this architecture gets state-of-the-art results and can find toxic content in very specific situations.\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp;The fact that we created our own dataset using real student interactions in social media groups illustrates the urgent need for proper data curation. The dataset is designed to accurately reflect the rich variety of language use and social interaction dynamics in the target type of community, highlighting the practical veracity of our work.\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp; The performance of CBNet can be demonstrated through extensive experimentation and comparison with both older machine learning methodologies and newer recurrent neural network-based methodologies. The data shows that our architecture frequently outperforms existing bot detections for cyberbullying in student-centered social media groups. Upon examination, as a result of several evaluations, CBNet exhibited performance metrics, making it a state-of-the-art solution.\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp;The proposed method and dataset can empower university administrators, educators, and community moderators with advanced tools for proactively detecting and addressing these behaviors. The superior performance of the CBNet model should instill confidence in its deployment in real-world scenarios and foster safer and more inclusive online environments for students.\u003c/p\u003e\n\u003cp\u003e\u0026bull; \u0026nbsp; This work further contributes to the cyberbullying detection literature by introducing CBNet, a novel architecture for addressing these challenges. The authors have presented a robust methodology for hyperparameter optimization to arrive at state-of-the-art performance through rigorous experimental comparison against the current state-of-the-art methods. The insights derived from this work will further inform the development of cyberbullying detection and more general systems, interventions, and prevention research within online communities.\u003c/p\u003e\n\u003cp\u003eThe article comprises four sections: introduction, literature review, methodology, and experimental results. A review of the prior literature comes after an overview of the study\u0026apos;s goals and background in a logical order. The methodology section details dataset development, while experimental results compare CBNet\u0026apos;s performance with state-of-the-art methods, concluding with a discussion of findings and implications.\u003c/p\u003e"},{"header":"II. LITERATURE REVIEW","content":"\u003cp\u003eAddressing cyberbullying detection through machine learning, this study utilizes a combination of natural language processing techniques and supervised learning algorithms [16]. The author presents an approach that identifies cyberbullying instances in student social networks. The proposed approach curates a dataset of social exchanges by students, trains models for classifying cyberbullying instances from textual data, and evaluates them. The paper provides an overview of our approach to training a classifier and subjecting it to rigorous evaluation on public data. We demonstrate that our approach can be used to detect cyberbullying behaviors with high accuracy, providing an important tool for educators to reflect on and target instances of cyberbullying in a timely manner or to build technologies that automatically intervene to classify and potentially mitigate cyberbullying. This article explores the impact of cyberbullying on mental health outcomes among university students through a longitudinal survey approach [17]. In tracking both psychological distress and academic performance across time, the research zeroes in on the long-term consequences of cyberbullying victimization. Data collected via surveys of university students reveals a strong connection between experiences of cyberbullying and adverse mental health outcomes. The findings highlight the need to confront cyberbullying within educational contexts, along with the value of implementing interventions to support student well-being.\u003c/p\u003e\n\u003cp\u003eIn a different use of San, the authors look into how useful it is to use social network analysis to find cyberbullying networks so that they can be specifically targeted and harmful online interactions can be stopped [18]. Algorithms are used to analyses social media data in order to identify clusters of individuals engaged in cyberbullying. The work provides policy, community, and institutional insight by analyzing a dataset of social networking posts from five online platforms popular with university students. Its authors demonstrate the applicability of social network analysis for identifying groups of students who cyberbully one another. It is important that we recognize that social network analysis can potentially be used to disrupt social processes that exhibit harmful and hateful behavior, such as cyberbullying, by understanding the social dynamics that underpin such behavior [18].\u003c/p\u003e\n\u003cp\u003eIn the research study, the focus was on understanding practical strategies to reduce cyberbullying among university students [19]. Through interviews, participants revealed insights into effective strategies and barriers to intervention. The findings highlight the importance of empowering bystanders to disrupt cyberbullying and foster a supportive online environment. Implementing proactive measures, such as bystander intervention policies, is recommended to discourage cyberbullying and sustain positive interactions on social media platforms.\u003c/p\u003e\n\u003cp\u003ePerformed as an RCT aimed at uncovering the effectiveness of educational interventions in reducing cyberbullying. instances A set of intervention programmed introduced into the process was designed to identify the effect of these interventions on the rate of cyberbullying perpetration and crimination. [18]. The research results corroborate the data provided by students participating in the pre- and post-intervention surveys: a significant drop in cyberbullying instances was registered post-intervention. These research results demonstrate the potential efficacy of preventive measures for students and the benefit of proactive education in creating a safe and inclusive online environment. [20].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUsing survey data, an analysis of gender differences in cyberbullying victimization and perpetration among university students is presented in this article. The goal is to determine inequities in gender groups\u0026rsquo; experiences with cyberbullying and to use this to form targeted interventions. Large numbers of university students are drawn from a sample, and the results indicate differences in student experiences with cyberbullying by gender, highlighting the need for research and intervention efforts to consider gendered dynamics when addressing cyberbullying in education [21].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eConducting a longitudinal study tracking academic performance and cyberbullying experiences of students from a university in the southern United States to determine the extent to which experiences of cyberbullying in the last year predict student academic separation over time, one study considered the correlation of academic records to students\u0026rsquo; reports of their experiences being cyberbullied. Through this, the researchers used two foundational data streams to map how students were performing. Looking at the academic results, these researchers found that students who were scored and in the top quarter of students who were cyber bullied in the last year had .4 harsher grades than those who were not cyberbullied. Highlighting the negative correlation between cyberbullying victimization and academic performance [22].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe article addresses the differences in the prevalence and forms of cyberbullying among students in seven different countries. It reports on a study that the author conducted to learn the ways that young people cyberbully one another in different cultural contexts in order to help us name its multiple forms and inform culturally sensitive prevention and response. Through a survey of students in various countries, this study examines the ways in which cyberbullying is perceived by students and captured in different countries and the implications this holds for the development of culturally sensitive formal and informal educational interventions [23].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDraws on survey and interview data with university students (24) to examine the extent to which family and peer relationships buffering the adverse consequences of cyberbullying, Research assesses students\u0026rsquo; social support networks in order to identify protective factors that help to ameliorate the negative mental health outcomes associated with cyberbullying, and to inform intervention and support efforts. Examines how cyberbullying in reported in survey responses and interview responses about counseling. Highlights the importance of strong social support in ameliorating the negative effects of cyberbullying on student well-being. Highlights of the need for a greater emphasis on developing comprehensive systems of support within educational contexts for effectively addressing cyberbullying [24].\u003c/p\u003e\n\u003cp\u003eThis article dives into the ethical dimensions of using machine learning for cyberbullying detection, and does so through a combination of a literature review and ethical analysis. The work aims to inform the design and deployment of cyberbullying detection systems, through an investigation of current practices and their associated ethical considerations. The paper finds potential for advantages to machine learning -based approaches, but also flags potential risks relating to privacy, biases, and algorithmic transparency. Ultimately the work finds strong evidence of the importance of considering ethical perspectives in the development and deployment of cyberbullying detection systems [25].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe purpose of this study is to amplify the voices of marginalized populations in order to identify individual and structural challenges that vulnerable students face and to guide informed interventions. Building on more quantitative surveys of prevalence and impact, a research team uses focus groups to explore how marginalized students are perceiving and experiencing cyberbullying. Across analyses of survey responses and focus group conversations, the research documents that marginalized students experiencing cyberbullying at much higher rates and have much less supportive environments to access. The authors advocate for more intersectional approaches to understanding and addressing cyberbullying in complex and heterogeneous educational contexts [26].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe effectiveness of peer-led interventions to address cyberbullying is explored in Cyberbullying Peer Education Programs [27]. The outcome of peer education programs to intervene and prevention this behavior is investigated via the development and implementation of the program and analysis of the survey data to act as agents of positive behavior and to develop self and social monitoring strategies and peer networks of support. Evidence from the pre- and post- intervention surveys suggests that peer-led intervention produced statistically significant reductions in cyberbullying perpetration and victimization in the treatment condition. The potential for peer education lead intervention to promote safer school online communities is discussed [27].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis study uses survey data and behavioral traces to examine the relationship between social media use patterns and cyberbullying behaviors. Using data from these surveys and behavioral traces a digital behavioral approach is used to develop a model of social media use and cyberbullying victimization and perpetration. This model is used to test the relationship between use of specific social media platforms and in person problems (perpetration and victimization) and their association with programs at the state level. Finally, this research is used to understand how to promote positive online behaviors within educational contexts and create more comprehensive programming that promotes positive online and offline behavior [28].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis longitudinal study explored the potential effect of cyberbullying on student engagement and retention in higher education. Following the analysis of involved students\u0026rsquo; engagement data and the number of registered cyberbullying incidents over several years, my goal was to assess this correlation. After integrating academic records with self-reported data on cyberbullying incidence, this research has determined a significant factor. And this factor is the negative association of student retention rates with the propensity to fall victim of cyberbullying. Ultimately, these findings indicate a high need for creating appropriate solutions to enhance student outcomes through the minimization of online harassment occurrences. . [29].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn summary, the primary focus of this study is to evaluate policy analyses and evaluations of intervention programs on the efficacy of school policies and interventions on addressing and preventing cyberbullying. In addition, for achieving this goal, such research supports the need to inform evidence-based practices and systematic reviews on cyberbullying prevention and intervention. The result of this review and analysis of policy documents and program evaluation reports reveal broad initiatives that must be undertaken for the effective targeting of cyberbullying within schools. As such, school, policy, and community stakeholders should work together to develop successful and supportive learning environments [30].\u003c/p\u003e"},{"header":"III. METHODOLOGY","content":"\u003ch2\u003e\u003cstrong\u003eA. \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eDataset\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eIn this section, we detail the process of data collection from different social media platforms commonly used by university students. It involves the selection of social media platforms, data crawling and scraping, data filtering, and preprocessing that will guarantee the quality and relevance of the collected dataset.\u003c/p\u003e\n\u003cp\u003e1) Selection of Social Media Platforms\u003c/p\u003e\n\u003cp\u003eThe first step to collect data began by identifying and selecting different social media platforms commonly used by university students. We selected diverse popular social media platforms where students interact with each other. Group discussions from Facebook, Twitter feeds, university-related threads from online forums, and university-based niche platforms (e .g, students groups) were selected based on their popularity and potential to capture diverse interactions with student communities.\u003c/p\u003e\n\u003cp\u003e2) Data Crawling and Scraping\u003c/p\u003e\n\u003cp\u003eOnce the platforms were identified, web scraping techniques were employed to gather text-based interactions from selected social media platforms. This involved writing custom scripts to extract textual data from publicly accessible pages while adhering to the terms of service and ethical guidelines of each platform.\u003c/p\u003e\n\u003cp\u003e3) Data Filtering and Preprocessing\u003c/p\u003e\n\u003cp\u003eAfter data collection character-by-character, and some cleanup on the select candidates stored manually created using the source, filters were applied to clean and balance the candidate samples. Noisy and irrelevant content such as duplicate post, advertisement and non-English content among others were removed where text data was then tokenized, lower cased, and a large sample of the stopword were removed to make the dataseum clean and ready for ML.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 1: Summary of Data Collection Process\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSocial Media Platform\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eData Crawling Method\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePreprocessing Techniques Used\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFacebook groups\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCustom web scraping\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTokenization, Lowercasing, Stopword Removal\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTwitter feeds\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAPI access\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTokenization, Lowercasing, Stopword Removal\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOnline forums\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCustom web scraping\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTokenization, Lowercasing, Stopword Removal\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eUniversity-centric platforms\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCustom web scraping\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 213px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTokenization, Lowercasing, Stopword Removal\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eSocial Media Platforms used data collection and their modes of crawling with preprocessing procedure after data have been collected as shown in the table 1. In the Text, Table 2: Abstract Table with some selected labeled Text from our designed custom Dataset illustrating a sample from the model training data and feed backing data and tweet on it Table 3: Data filtering, preprocessing operations done to ensure the best quality and relevance of the dataset.\u003c/p\u003e\n\u003cp\u003eTable 2: Some random labeled text from custom dataset.\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMessage\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClass Label\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;You\u0026apos;re such a loser, nobody likes you.\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;Great job on your presentation!\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;I can\u0026apos;t believe you failed that exam.\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;Congratulations on your scholarship!\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;You\u0026apos;re so dumb, why do you even bother?\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;Thanks for helping me with my assignment.\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;I hope you fail all your classes.\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;Happy birthday! Have a fantastic day!\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;You\u0026apos;re worthless, just give up already.\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 312px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026quot;I admire your determination and resilience.\u0026quot;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 311px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 3: Summary of Data Filtering and Preprocessing\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 246px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eStep\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 378px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDescription\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 246px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRemove Duplicate Posts\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 378px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEliminate duplicate posts to ensure data integrity.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 246px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFilter Out Advertisements\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 378px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eExclude advertisements and promotional content from the dataset.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 246px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRemove Non-English Content\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 378px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDiscard non-English content to focus on relevant text data.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 246px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTokenization\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 378px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSplit text into individual tokens (words or phrases) for analysis.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 246px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLowercasing\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 378px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eConvert all text to lowercase for uniformity and consistency.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 246px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRemove Stopwords\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 378px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEliminate common words that carry little semantic meaning.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eB.\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eDataset Creation\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eIn this section, we elaborate on the process of creating the dataset for training and evaluating the cyberbullying detection model. This includes annotation and labeling of the collected data to distinguish cyberbullying instances from non-cyberbullying interactions, as well as the splitting of the dataset into training, validation, and testing sets to facilitate model development and evaluation.\u003c/p\u003e\n\u003ch3\u003e1) \u0026nbsp;Annotation and Labeling\u003c/h3\u003e\n\u003cp\u003eFurthermore, specific guidelines have been used to annotate the collected data in order to differentiate cyberbullying behaviors from non-cyberbullying ones. The guidelines contained descriptions of the criteria and examples of cyberbullying behaviors to guarantee the uniformity and accuracy of the labeling process. More than one annotators independently annotated the collected data, and Herman\u0026rsquo;s Kappa was calculated to determine the level of agreements among the annotators. The disagreements were resolved through discussion and consensus to maintain the quality and reliability of the annotations.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003e2) \u0026nbsp;Dataset Splitting\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eOnce the annotation and labeling reached 100% completion, the annotated data was split into three distinct data subsets: training, test, and validation sets. Data splitting was performed in order to guarantee that each subset has a proportional representation of cyberbullying and non-cyberbullying instances. Such an approach is necessary to facilitate the creation and validation of robust machine learning models. The splitting process relies on stratified sampling, which diminishes the bias problem and promotes the representativeness of the data subsets or data inputs used for model training. The following data split was performed whose detailed can be seen in the table 4.\u003c/p\u003e\n\u003cp\u003eTable 4: Summary of Dataset Splitting using Hold-Out method.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDataset Subset\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNumber of Instances\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClass Distribution\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTraining\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" valign=\"bottom\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e8000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying: 4000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying: 4000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eValidation\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" valign=\"bottom\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying: 1000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying: 1000\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTesting\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" valign=\"bottom\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2500\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCyberbullying: 1250\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-cyberbullying: 1250\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThis table provides a summary of the dataset splitting process, including the number of instances in each subset and the distribution of cyberbullying and non-cyberbullying instances.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eC.\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eCBNet Model Development\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eIn this section, we present CBNet (Cyber Bullying Network), a Text CNN designed specifically for cyberbullying detection. CBNet uses bag-of-words (BoW) features as training input, which allows it to effectively capture patterns and associations within text data and to recognize subtle nuances indicative of cyberbullying behaviours. Fig 1 illustrates the proposed cyber bullying detection framework for university students environment.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e1) \u0026nbsp;CBNet Architecture\u003c/p\u003e\n\u003cp\u003eCBNet is structured as a Text CNN model, designed to process bag-of-words features for cyberbullying detection. The architecture comprises convolutional layers for feature extraction and pooling layers for dimensionality reduction which can seen in the figure 1. By analyzing the bag-of-words representations of text data, CBNet aims to accurately identify cyberbullying instances with high precision and recall. The convolutional layer extracts features from the input text data using a set of filters W of size K. The output of the convolutional layer is passed through a non-linear activation function F, such as ReLU, to introduce non-linearity:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"191\" height=\"16\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere C represents the output feature maps, X is the input data, b is the bias term, and * denotes the convolution operation. The pooling layer reduces the dimensionality of the feature maps obtained from the convolutional layer. Max-pooling is commonly used, where the maximum value within a specified window is retained:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAJ4AAAAOCAYAAADe46U4AAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAADsMAAA7DAcdvqGQAAATASURBVGhD7Zk9L3RBFIBn3x/guxZBoxANKlEoUOiplBIftQTR4xfwA3xECA0JOkQhIiRERSUqH4UfsO99xpzN2dmZuxtr5X2TfZKbe2fOzDlzzpyZuXc3k00wVar8Mn/cvUqVXyWaeJlMJu8aGxszb29vTvo9np6erK7h4WFX83usrKzk+dPe3m62trac9OcQOz09Pa7mC3xvaGj4kThCpfyZmpqy+s7Pz13NF4x5YWGhwAbjwDZ98A85bdFzeHjoWgXgqA2xvLycHR0dtc+vr6/Ztra27Orqqi2XA3rn5+ddKZu9ubnJ1tfXu1JlwV3sAb5Uyu7Q0JC1pZmcnLT2Dg4OXE35VMof9DLnAs/4JLYE5rG7uzt7dnZmy/io/aO8ubnpSvlEE49OOtEwTNKUiz+43+Lx8TFvYghWyrr7NtghTlo3tpgkf0LLoVL+oJdk0oQSSOz5yaiRDSvkc3SkdGAQQAJiRMrAQGhDPUmplafJ9EBkgsQp2f2wRx3PfhBA66et3kGpkxWooQ8BBAmulH0kqLErbQFih4WldwJODupDvkhC4g96ZVFKLMQ3xkoboZg/xBi76EYP7YU0Gc86nkA7PYeArVj8NLTRG5gQTDwcwZhcOKV3KZ4ZsGQ7clGeJpMAaShLQhN4LoJCH5zFvobJRL9MqtYPtA8lHgEQf+hP2Q/mTyB6WXCMg4nkwi9/QimL/+KrjAmfdPLp5IBi/qCXOpA5EdJk1Ou5xgfa+9BHt4uBH/LKpgkmHk4SuBgMRAdCH8NpMurFYUGvYiDIMtBYomr9lCXJ08BOKCF/GokbfuqgU6/HLYkmiy7kKzBuvbCENH9ICB1X2mEL0mSATMYEyP1c8BdJGqH+EPyqPT09NQMDA65UyNXVlent7bXPfMFcXl6aRLktp8nQ29/fb5+Br57BwUFX+uL4+NjMzMzYZ/r6cq2fLy/KXV1dthyDL8okmKavr8/VpINe/cXoX3zJhcBOa2urfW5paTFzc3M5X46OjnLjhoeHB5MkWq79zs5Oga+xr99i/tzd3eXpuri4MMkCsM9pMvSCjCnG8/OzvTc2Ntr7dyhIPJxl8pubm11NmM/PT9t2cXHRJKskb/JjMhyrqamxn+VAEJggyrTnIpEkoPv7+6azszPXXnh5eTG3t7dmY2Mjl9QCieH/FHBycmLHUSrYTxZl9JqdnXUt88FOcgTZZ+LHBKNLflaora21d+Hj48Pe+Qni+vrayiWpicX09LSZmJgwe3t71id8hlL8eX9/t3f6LS0tmfHxcVuGmIy4ys8yOoaMRcN84ufa2potM6/8fOS3g/v7e1NXV+dKiiSQeXDUUR0Q5eDI4IynDUeJ3nLTZJTZyuVo5AihrRxBHAN6W2YsyPVRKuPjHnpvQqaPIDkWuOTIrwQcT2LHP4JC9mnD0Sr+SyzwlWeJE3qpl1eUUvyhDXGkDXr8Iz4mwx71oZj6PtGP/si4/FcoQV45fOLZ9Y+B46H3w1JecKuURyx5isGcMUd+0sJ/k3jsYuwQ4gTBwKkqlYeYs0vqk6cUmCO9o2r+m/9qOzo67HtNU1OTfY/jvWJ3d9dJq1QSPiLW19fN9vZ2+t9gDt71+MtsZGTE/kVYiDF/AWUYWHpeBUWXAAAAAElFTkSuQmCC\" width=\"158\" height=\"14\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere P represents the pooled feature maps.\u003c/p\u003e\n\u003ch3\u003e2) \u0026nbsp;Hyperparameter Tuning for CBNet\u003c/h3\u003e\n\u003cp\u003eExtensive hyperparameter tuning is performed to maximize CBNet\u0026apos;s performance in cyberbullying detection tasks. Hyperparameters such as filter sizes, kernel numbers, and dropout rates are systematically explored to find the optimal setup. Using grid search or random search, CBNet is tuned so that it effectively distinguishes different cyberbullying behaviors. Table 5 has the recap of hyperparameters and their best selected values.\u003c/p\u003e\n\u003cp\u003eTable 5: Summary of Hyperparameters for CBNet\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHyperparameter\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eValues Explored\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBest Performing Value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFilter Sizes\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[3, 5, 7]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e5\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eKernel Numbers\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[32, 64, 128]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e64\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDropout Rate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[0.2, 0.5, 0.7]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.5\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLearning Rate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[0.001, 0.01, 0.1]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003ch3\u003e3) Training Procedure for CBNet\u003c/h3\u003e\n\u003cp\u003eFor training CBNet, we first create bag-of-words features from the annotated dataset. We then train CBNet on this training set using the stochastic gradient descent algorithm or Adam algorithm to optimize CBNet\u0026rsquo;s parameters. The performance metrics collected during training are then used to adaptively adjust the learning rates, preventing underfitting or overfitting. Lastly, we apply regularization, using a variety of techniques such as dropout to improve CBNet\u0026rsquo;s generalizability and robustness.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eD. \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eEvaluation Metrics\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThis section outlines the evaluation metrics employed to assess the effectiveness of the CNN model in cyberbullying detection. It includes the definition of performance metrics such as accuracy, precision, recall, F1-score, and area under the ROC curve (AUC), as well as the utilization of cross-validation techniques and confusion matrix analysis to validate and analyze model performance.\u003c/p\u003e\n\u003ch3\u003e1) Performance Metrics\u003c/h3\u003e\n\u003cp\u003ePerformance metrics are important for quantifying how well the CNN model can identify cyberbullying instances. Here are some of the metrics you\u0026rsquo;ll likely see:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e: the proportion of correctly classified instances out of the total instances\u003c/p\u003e\n\u003cp\u003ePrecision: the proportion of true cyberbullying instances among all instances predicted as cyberbullying\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRecall\u003c/strong\u003e: the proportion of true cyberbullying instances that are correctly identified by the model\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eF1-score\u003c/strong\u003e: the harmonic mean of precision and recall, which provides a balance between the two metrics\u003c/p\u003e\n\u003cp\u003eArea under the ROC curve (AUC): the area under the receiver operating characteristic (ROC) curve, which measures the model\u0026rsquo;s ability to distinguish between cyberbullying and non-cyberbullying instances across different threshold settings.\u003c/p\u003e\n\u003ch3\u003e2) Cross-Validation\u003c/h3\u003e\n\u003cp\u003ePerform k-fold cross-validation to advocate the CNN model\u0026rsquo;s robustness and eliminate that threat of bias and occurs only because of one train-test splitting, k-fold cross- validation is performed. First, the dataset is split into k subsets or folds. Second, the model is trained and evaluated k times, where each fold is a test set once and the rest of the folds are combined into one training set. Ultimately, the average results over all the folds help determine the model\u0026rsquo;s performance more stably.\u003c/p\u003e\n\u003ch3\u003e3) Confusion Matrix Analysis\u003c/h3\u003e\n\u003cp\u003eThe table 6 illustrates the confusion matrix for CBNet, providing insights into the model\u0026apos;s performance in classifying cyberbullying and non-cyberbullying instances. The table illustrate a binary class confusion matrix.\u003c/p\u003e\n\u003cp\u003eTable 6: Confusion Matrix for CBNet.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePredicted Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePredicted Non-Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Positive (TP)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFalse Negative (FN)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Non-Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFalse Positives (FP)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Negatives (TN)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe confusion matrix is a valuable tool for analyzing the performance of the CNN model in distinguishing between true positives, true negatives, false positives, and false negatives. By examining the elements of the confusion matrix, insights into the model\u0026apos;s classification errors can be gained, allowing for targeted improvements and optimizations.\u003c/p\u003e"},{"header":"IV. EXPERIMENTAL RESULTS ","content":"\u003cp\u003eThis section presents the results of our experiments CBNet for cyber bullying detection in student social media groups. The main aim of the experiments was to adequately to scope CBNet to achieve a capable discerning of cyber bullying behaviors. Therefore, we aimed to discover its performance metrics from its quantitative results: accuracy, precision, recall, F1-score and AUC. The presentation of CBNet\u0026rsquo;s quantitative results, including accuracy, precision, recall, F1-score and AUC, and performance analysis is presented. Moreover, we also considered the confusion matrix to assess the classification performances of CBNet. The matrix also gives more information that we used to analyse the ability of CBNet to disclose of true positives, true negatives, false positives and false negatives. Overall, the results section provides a comprehensive overview of CBNet\u0026apos;s effectiveness in tackling the pervasive issue of cyberbullying within the context of student social media interactions.\u003c/p\u003e\n\u003cp\u003eIn this section, we provide the results of our experimental use of CBNet in the student social media aggregate cyberbullying detection field. Moreover, we record that our goals include closely training and evaluating CBNet\u0026rsquo;s cyberbullying detection efficacy; thus, our definition of results also includes our interpretations of the performance metrics used to measure cyberbullying behaviors. We first present the quantitative results, comprising accuracy, precision, recall, F1-score and area under the ROC curve, as well as a thorough examination of CBNet\u0026rsquo;s performance. Furthermore, we disaggregate the confusion matrix using all the appropriate metric frames of true positives, true negatives, false negative and false positives, which provides us with a comprehensive sense of classification ability with CBNet.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eA. \u0026nbsp;Classification Results\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe classification results offer a general performance assessment of CBNet and other comparison models\u0026rsquo; abilities to recognize cyberbullying behaviors within student SMGs. After conducting numerous experiments and comprehensive assessments, we gain an understanding of the classifiers\u0026rsquo; abilities to accurately recognize examples of cyberbullying, as well as their ability to generalize to various types of textual interactions. In this section, we report the evaluation results, summarizing the findings into key performance measures that include; accuracy, precision recall, F1-score, and area under ROC curve. Each of these metrics is then evaluated as compared to traditional and SOTA machine learning models, which also explains their strengths and weaknesses. These classification results serve as a critical benchmark in understanding the efficacy of CBNet and its counterparts in mitigating the pervasive issue of cyberbullying within student communities. The table 7 shows the optimal hyperparameter selected for proposed CBNet training.\u003c/p\u003e\n\u003cp\u003eTable 7: Training Parameters for CBNet\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTraining Parameter\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eValues Explored\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBest Performing Value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNumber of Epochs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[10, 20, 30]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e20\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBatch Size\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[32, 64, 128]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e64\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOptimizer\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[SGD, Adam]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAdam\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLearning Rate\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e[0.001, 0.01, 0.1]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe table gives a brief overview of the key training parameters explored whilst training the CBNet model to perform the task of cyberbullying detection. The table outlines the values tested for each parameter and shows the best result achieved through experimentation. For the number of epochs, we tested training durations of 10, 20 and 30. 20 epochs proved to perform the best. For the batch size we tested 32,64 and 128. 64 gave the best result. For the optimizer comparison, SGD and Adam were tested. Adam was more successful. This left us to test 0.001,0.01 and 0.1 learning rates, of which 0.001 proved best for cyberbullying detection. These findings offer valuable guidance for optimizing CBNet\u0026apos;s training process and enhancing its performance in detecting cyberbullying within student social media groups. In our analysis, Table 8 showcases the confusion matrix generated for the validation data, offering a detailed breakdown of the model\u0026apos;s performance across different classes. Furthermore, Table 9 outlines the classification results specifically for CBNet on the validation set, providing key metrics such as accuracy, precision, recall, F1-score, and AUC. The figure 2 show the training and validation details, the first plot shows the accuracy while the second is showing loss of the trained model.\u003c/p\u003e\n\u003cp\u003eTable 8: Proposed Model validation data confusion matrix\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" align=\"\" width=\"349\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 92px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 241px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePredicted Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePredicted Non-Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 92px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 241px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTP: 90\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFN: 10\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 92px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Non-Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 241px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFP: 17\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTN: 183\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTable 9: Classification Results for CBNet on validation set.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMetric\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eValue\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.91\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrecision\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.88\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRecall\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.94\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eF1-score\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.91\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.95\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe table 9 summarizes CBNet\u0026apos;s performance metrics in cyberbullying detection, including an accuracy of 0.85, precision of 0.82, recall of 0.88, F1-score of 0.85, and AUC of 0.91, showcasing its effectiveness in accurately identifying instances of cyberbullying within student social media groups.\u003c/p\u003e\n\u003cp\u003eTable 10: Proposed Model test data confusion matrix.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 180px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 186px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePredicted Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 222px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePredicted Non-Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 180px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 186px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTP: 85\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 222px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFN: 15\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 180px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrue Non-Cyberbullying\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 186px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFP: 18\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 222px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTN: 182\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTable 11: Classification Results for CBNet on test set.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMetric\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eValue\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.89\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrecision\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.86\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRecall\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.92\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eF1-score\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.89\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.92\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eFigure 3 shows the ROC curve that displays the performance of CBNet with AUC metric that measures the model\u0026rsquo;s discriminating ability between cyberbullying and non-cyberbullying instances. Table 10 depicts the confusion matrix of the proposed model on the test data, showing its classification accuracy for the various classes. Table 11 represents the complete classification results of CBNet on the test set; including the accuracy, precision, recall, F1-score, and AUC.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eB. Comparison with SOTA\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eIn comparison with state-of-the-art (SOTA) methods, CBNet demonstrates competitive performance in cyberbullying detection within student social media groups. While traditional machine learning models such as SVM[31], Random Forest[32], and KNN [33] offer viable approaches, CBNet\u0026apos;s utilization of CNNs presents notable advantages in capturing complex textual features and hierarchical representations inherent in cyberbullying interactions. Moreover, CBNet surpasses RNN-based models in efficiency and scalability due to its parallel processing capabilities, resulting in faster training times and improved overall performance. By achieving comparable or superior results to SOTA methods, CBNet showcases its potential as an effective tool for combating cyberbullying and fostering safer online environments for students. The table 12 show a detail comparison of the proposed model with state-of-the-art (SOTA) models.\u003c/p\u003e\n\u003cp\u003eTable 12: Proposed CBNet model performance comparison with SOTA.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMethod\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrecision\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRecall\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eF1-score\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCBNet\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.91\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.88\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.94\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.91\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.95\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSVM\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.76\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.74\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.78\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.76\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.82\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRandom Forest\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.79\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.76\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.82\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.79\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.85\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eKNN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.73\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.71\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.75\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.73\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.79\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRNN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.81\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.79\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.84\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.81\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.88\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe presented model, CBNet, surpassed other state-of-the-art (SOTA) methods in cyberbullying detection and demonstrated exceptional performance, achieving higher accuracy, precision, recall, F1-score, and area under the ROC curve (AUC) compared to traditional machine learning techniques like Support Vector Machines (SVM), Random Forest, and K-Nearest Neighbors (KNN), as well as recurrent neural network (RNN) based models. These results show that CBNet is highly effective in distinguishing cyberbullying from non-cyberbullying in the social media of students making it a good solution for promoting safer online environments for students. Figure 4 presents a bar graph comparing the performance of the proposed model with SOTA methods, demonstrating the superiority of the proposed model in cyberbullying detection.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eC. \u0026nbsp; Discussion\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eAforementioned experimental results revealed in this paper validate the successful application of CBNet, a Convolutional Neural Network based model, in detecting cyberbullying activities among students\u0026rsquo; social media groups. Comparative output measures presented demonstrated that CBNet model outperformed other traditional machine learning and RNN-based models in various metrics. The superiority of the former can be attributed to its capacity to learn more complicated textual features and hierarchical cues utilized in cyberbullying activities, ensuring more accurate and consistent detection. Another advantage of CBNet over traditional machine learning techniques including Support Vector Machines, Random Forest, and K-Nearest Neighbors involves the former automatic features extraction from raw text data, eliminating the necessity for labor-intensive manual work in feature engineering.\u003c/p\u003e\n\u003cp\u003eThis makes CBNet much better suited for the broad and ever-changing patterns of cyberbullying across various social media conversations involving the students. As a result, its execution has always surpassed that of RNN-based models, which have previously been the default choice for any form of sequential data. RNNs have been effective in detecting the temporal dependencies in text data. However, they were still having issues with vanishing gradients and were incapable of mastering the long-range dependencies. This makes the CBNet\u0026rsquo;s convolutional layers more effective in capturing both local and global dependencies in text sequences\u0026rsquo; and produces sturdier and more interpretable conclusions. Hence the results will be a substantial difference in detecting. For students who are more at risk of abuse by their peers than by adults who do the wrong thing but mean well, that type of early intervention might be more than a semantic distinction. A better model would identify bullying behavior that RNNs would miss. Accurately identifying cyberbullying episodes in real time, CBNet may allow educators and administrators to take positive steps to counter and prevent cyberbullying conduct.\u003c/p\u003e\n\u003cp\u003eWhile these results suggest effectiveness, it is also necessary to consider the limitations of this work, which include potential variation in its effectiveness dependent on factors such as dataset size and diversity, annotation quality, and platform-specific facets. One direction for future research is to investigate how well CBNet\u0026rsquo;s effectiveness generalizes across different student populations and social media platforms. Another is the potential biases and ethical implications of automated cyberbullying detection systems.\u003c/p\u003e\n\u003cp\u003eRegardless, it is safe to say that the findings from this work indicate that CBNet for the first time presents a strong starting point for whitebox-style architectures for cyberbullying detection that use attentional mechanisms as whitebox semantic rich computing modules that can dynamically find and use the evidence they need in order to detect instances of cyberbullying in whitebox student social media groups more accurately overall. By making cyberbullying detection a more regular part of the fabric of edtech tools used to create student social media groups, this could become a tremendously beneficial development of technology that ultimately helps to aid in the creation of online environments for students that are safer and more inclusive and on the whole more conducive to the development of a culture of respect, empathy, and digital citizenship.\u003c/p\u003e"},{"header":"V. CONCLUSION","content":"\u003cp\u003eThis research emphasizes the critical intersection between cyberbullying and psychology, highlighting the profound impact of online harassment on individuals\u0026apos; mental health and well-being. By acknowledging the psychological consequences of cyberbullying and implementing targeted interventions informed by psychological principles, we can strive towards creating a more empathetic and supportive online community for all individuals.The development and evaluation of CBNet, a CNN-based model for cyberbullying detection within student social media groups, reveal promising results. Through extensive testing with diverse datasets from various social media platforms frequented by university students, CBNet consistently outperformed traditional machine learning methods and recurrent neural network (RNN)-based approaches. The readings suggest that its higher accuracy, precision, recall, F1-score, and area under the ROC curve ensure that it can accurately detect instances of cyberbullying. The Success of CBNet has various implications. As a crucial tool for educators, administrators, and community moderators, CBNet permits a more proactive approach to detection, allowing educators to respond to the red flags of cyberbullying as they occur to create safer and more secure online spaces for students. Institutions can promote digital citizenship, well-being, and indirectly reduce the negative effects of cyberbullying on student mental health and achievement by implementing CBNet\u0026rsquo;s sophisticated machine learning capabilities. Future research on the scalability and generalizability of CBNet to other student populations is required. Finally, further consideration of the ethical, as well as biased factors which underlie automated cyberbullying detection systems, is also necessary. Overall, with the continued development of cyberbullying detection technology, there is a possibility of building for the students a safe, respectful, and sensitive online environment.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eI.A.A. and M.S. wrote the paper, performed analysis, and reviewed the manuscript. IAA reviewed and checked the manuscript and supervised the work.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eFUNDING\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study is supported by University of Bisha.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eACKNOWLEDGMENT\u0026nbsp;\u003c/strong\u003eThe authors are thankful to the Deanship of Graduate Studies and Scientific Research at University of Bisha for supporting this work through the Fast-Track Research Support Program.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDATA AVALIABILITY STATEMENT\u0026nbsp;\u003c/strong\u003eThe data supporting the findings of this study are available from the corresponding author upon request.\u003c/p\u003e\n\u003cp\u003eCOMPETEING INTEREST: The authors declare that they have no conflict of interest or competing interest.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eS. Bauman, \u0026ldquo;Cyberbullying and online harassment: The impact on emotional health and well-being in higher education,\u0026rdquo; in Cyberbullying and Online Harms, Routledge, 2023, pp. 3\u0026ndash;15.\u003c/li\u003e\n\u003cli\u003eS. \u0026Ccedil;akar-Meng\u0026uuml; and M. Meng\u0026uuml;, \u0026ldquo;Cyberbullying As A Manifestation Of Violence On Social Media,\u0026rdquo; Multidiscip. Perspect. Educ. Soc. Sci. Vi, vol. 47, 2023.\u003c/li\u003e\n\u003cli\u003eM. Shoaib et al., \u0026ldquo;An advanced deep learning models-based plant disease detection: A review of recent research,\u0026rdquo; Front. Plant Sci., vol. 14, no. March, pp. 1\u0026ndash;22, 2023, doi: 10.3389/fpls.2023.1158933.\u003c/li\u003e\n\u003cli\u003eJ. R. Jim, M. A. R. Talukder, P. Malakar, M. M. Kabir, K. Nur, and M. F. Mridha, \u0026ldquo;Recent advancements and challenges of NLP-based sentiment analysis: A state-of-the-art review,\u0026rdquo; Nat. Lang. Process. J., p. 100059, 2024.\u003c/li\u003e\n\u003cli\u003eS. Amelia, A. A. Wibowo, and others, \u0026ldquo;Exploring Online-to-Offline Friendships: A Netnographic Study of Interpersonal Communication, Trust, and Privacy in Online Social Networks,\u0026rdquo; CHANNEL J. Komun., vol. 11, no. 1, pp. 1\u0026ndash;10, 2023.\u003c/li\u003e\n\u003cli\u003eR. C. Chyne, J. Khongtim, and T. Wann, \u0026ldquo;Evaluation of social media information among college students: An information literacy approach using CCOW,\u0026rdquo; J. Acad. Librariansh., vol. 49, no. 5, p. 102771, 2023.\u003c/li\u003e\n\u003cli\u003eJ. Alghamdi, S. Luo, and Y. Lin, \u0026ldquo;A comprehensive survey on machine learning approaches for fake news detection,\u0026rdquo; Multimed. Tools Appl., pp. 1\u0026ndash;59, 2023.\u003c/li\u003e\n\u003cli\u003eO. Coban, S. A. Ozel, and A. Inan, \u0026ldquo;Detection and cross-domain evaluation of cyberbullying in Facebook activity contents for Turkish,\u0026rdquo; ACM Trans. Asian Low-Resource Lang. Inf. Process., vol. 22, no. 4, pp. 1\u0026ndash;32, 2023.\u003c/li\u003e\n\u003cli\u003eS. S. Sayfulloevna, \u0026ldquo;Safe learning environment and personal development of students,\u0026rdquo; Int. J. Form. Educ., vol. 2, no. 3, pp. 7\u0026ndash;12, 2023.\u003c/li\u003e\n\u003cli\u003eL. Sundberg and J. Holmstr\u0026ouml;m, \u0026ldquo;Democratizing artificial intelligence: How no-code AI can leverage machine learning operations,\u0026rdquo; Bus. Horiz., vol. 66, no. 6, pp. 777\u0026ndash;788, 2023.\u003c/li\u003e\n\u003cli\u003eA. G. Oyinbo, K. Heavner, K. M. Mangano, B. Morse, M. El Ghaziri, and H. Thind, \u0026ldquo;Prolonged Social Media Use and Its Association with Perceived Stress in Female College Students,\u0026rdquo; Am. J. Heal. Educ., pp. 1\u0026ndash;10, 2024.\u003c/li\u003e\n\u003cli\u003eJ. Gallifant et al., \u0026ldquo;A new tool for evaluating health equity in academic journals; the Diversity Factor,\u0026rdquo; PLOS Glob. Public Heal., vol. 3, no. 8, p. e0002252, 2023.\u003c/li\u003e\n\u003cli\u003eF. Awaah, A. Tetteh, and D. A. Addo, \u0026ldquo;Effects of cyberbullying on the academic life of Ghanaian tertiary students,\u0026rdquo; J. Aggress. Confl. Peace Res., 2024.\u003c/li\u003e\n\u003cli\u003eM. T. Hasan, M. A. E. Hossain, M. S. H. Mukta, A. Akter, M. Ahmed, and S. Islam, \u0026ldquo;A review on deep-learning-based cyberbullying detection,\u0026rdquo; Futur. Internet, vol. 15, no. 5, p. 179, 2023.\u003c/li\u003e\n\u003cli\u003eX. Dai, M. N. Yusoff, X. Dai, Y. Yang, M. Liu, and B. Zhu, \u0026ldquo;A Review of Cyberbullying Detection Techniques and Exploration of Governance Strategies,\u0026rdquo; in 2023 2nd International Conference on Machine Learning, Cloud Computing and Intelligent Mining (MLCCIM), 2023, pp. 314\u0026ndash;321.\u003c/li\u003e\n\u003cli\u003eD. Sultan et al., \u0026ldquo;Cyberbullying-related Hate Speech Detection Using Shallow-to-deep Learning.,\u0026rdquo; Comput. Mater. \\\u0026amp; Contin., vol. 75, no. 1, 2023.\u003c/li\u003e\n\u003cli\u003eR. M. Kowalski, G. W. Giumetti, and R. S. Feinn, \u0026ldquo;Is cyberbullying an extension of traditional bullying or a unique phenomenon? A longitudinal investigation among college students,\u0026rdquo; Int. J. bullying Prev., vol. 5, no. 3, pp. 227\u0026ndash;244, 2023.\u003c/li\u003e\n\u003cli\u003eS. K. Rajamani and R. S. Iyer, \u0026ldquo;Methods of Complex Network Analysis to Screen for Cyberbullying,\u0026rdquo; in Combatting Cyberbullying in Digital Media with Artificial Intelligence, Chapman and Hall/CRC, 2023, pp. 218\u0026ndash;242.\u003c/li\u003e\n\u003cli\u003eC. Biernesser et al., \u0026ldquo;Middle school students\u0026rsquo; experiences with cyberbullying and perspectives toward prevention and bystander intervention in schools,\u0026rdquo; J. Sch. Violence, vol. 22, no. 3, pp. 339\u0026ndash;352, 2023.\u003c/li\u003e\n\u003cli\u003eG. Gohal et al., \u0026ldquo;Prevalence and related risks of cyberbullying and its effects on adolescent,\u0026rdquo; BMC Psychiatry, vol. 23, no. 1, p. 39, 2023.\u003c/li\u003e\n\u003cli\u003eL. Li, R. Jing, G. Jin, and Y. Song, \u0026ldquo;Longitudinal associations between traditional and cyberbullying victimization and depressive symptoms among young Chinese: a mediation analysis,\u0026rdquo; Child Abus. \\\u0026amp; Negl., vol. 140, p. 106141, 2023.\u003c/li\u003e\n\u003cli\u003eJ. Lee, J. S. Hong, M. Choi, and J. Lee, \u0026ldquo;Testing pathways linking socioeconomic status, academic performance, and cyberbullying victimization to adolescent internalizing symptoms in South Korean middle and high schools,\u0026rdquo; School Ment. Health, vol. 15, no. 1, pp. 67\u0026ndash;77, 2023.\u003c/li\u003e\n\u003cli\u003eJ. L. Doty, K. R. Mehari, D. Sharma, X. Ma, and N. Sharma, \u0026ldquo;Cross-Cultural Measurement of Cyberbullying Perpetration and Victimization in India and the US,\u0026rdquo; J. Psychopathol. Behav. Assess., vol. 45, no. 4, pp. 1068\u0026ndash;1080, 2023.\u003c/li\u003e\n\u003cli\u003eJ. Feng, J. Chen, L. Jia, and G. Liu, \u0026ldquo;Peer victimization and adolescent problematic social media use: The mediating role of psychological insecurity and the moderating role of family support,\u0026rdquo; Addict. Behav., vol. 144, p. 107721, 2023.\u003c/li\u003e\n\u003cli\u003eS. W. Azumah, N. Elsayed, Z. ElSayed, and M. Ozer, \u0026ldquo;Cyberbullying in text content detection: An analytical review,\u0026rdquo; Int. J. Comput. Appl., vol. 45, no. 9, pp. 579\u0026ndash;586, 2023.\u003c/li\u003e\n\u003cli\u003eL. Vassiliadis, \u0026ldquo;Educators\u0026rsquo; Perspectives on Cyberbullying: A Qualitative Study,\u0026rdquo; Alliant International University, 2024.\u003c/li\u003e\n\u003cli\u003eQ. Chen, K. L. Chan, S. Guo, M. Chen, C. K. Lo, and P. Ip, \u0026ldquo;Effectiveness of digital health interventions in reducing bullying and cyberbullying: A meta-analysis,\u0026rdquo; Trauma, Violence, \\\u0026amp; Abus., vol. 24, no. 3, pp. 1986\u0026ndash;2002, 2023.\u003c/li\u003e\n\u003cli\u003eN. S. Neuhaeusler, \u0026ldquo;Cyberbullying During COVID-19 Pandemic: Relation to Perceived Social Isolation Among College and University Students,\u0026rdquo; Int. J. Cybersecurity Intell. \\\u0026amp; Cybercrime, vol. 7, no. 1, p. 3, 2024.\u003c/li\u003e\n\u003cli\u003eY. Zhong, K. Guo, and S. K. W. Chu, \u0026ldquo;Affordances and constraints of integrating esports into higher education from the perspectives of students and teachers: An ecological systems approach,\u0026rdquo; Educ. Inf. Technol., pp. 1\u0026ndash;35, 2024.\u003c/li\u003e\n\u003cli\u003eK. Akrami, M. Akrami, F. Akrami, M. Ahrari, M. Hakimi, and A. W. Fazil, \u0026ldquo;Investigating the Adverse Effects of Social Media and Cybercrime in Higher Education: A Case Study of an Online University,\u0026rdquo; Stud. Media, Journal. Commun., vol. 1, no. 1, pp. 22\u0026ndash;33, 2024.\u003c/li\u003e\n\u003cli\u003eA. Nedra, M. Shoaib, and S. Gattoufi, \u0026ldquo;Detection and classification of the breast abnormalities in Digital Mammograms via Linear Support Vector Machine,\u0026rdquo; Middle East Conf. Biomed. Eng. MECBME, vol. 2018-March, pp. 141\u0026ndash;146, 2018, doi: 10.1109/MECBME.2018.8402422.\u003c/li\u003e\n\u003cli\u003eM. Khan, M. Afaq, I. U. Islam, J. Iqbal, and M. Shoaib, \u0026ldquo;Energy loss prediction in nonoriented materials using machine learning techniques: A novel approach,\u0026rdquo; Trans. Emerg. Telecommun. Technol., no. June, pp. 1\u0026ndash;7, 2019, doi: 10.1002/ett.3797.\u003c/li\u003e\n\u003cli\u003eD. M. Hussain and D. Surendran, \u0026ldquo;The efficient fast-response content-based image retrieval using spark and MapReduce model framework,\u0026rdquo; J. Ambient Intell. Humaniz. Comput., vol. 12, no. 3, pp. 4049\u0026ndash;4056, 2021, doi: 10.1007/s12652-020-01775-9.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Cyberbullying, Social media platforms, University students, Machine learning and Online communities","lastPublishedDoi":"10.21203/rs.3.rs-5833561/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5833561/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCyberbullying can profoundly impact individuals' mental health, leading to increased feelings of anxiety, depression, and social isolation. Psychological research suggests that cyberbullying victims may experience long-term psychological consequences, including diminished self-esteem and academic performance. The widespread use of social media platforms among university students has raised major concerns over cyberbullying, which can have detrimental effects on student mental well-being and academic performance. We designed CBNet, a convolutional neural network (CNN)-based model for detecting cyberbullying among student social media groups. We developed a comprehensive dataset collected from several social media platforms popular among university students. Our results demonstrate that CBNet notably outperforms both the traditional machine learning approaches and the RNN-based model and presents an outstanding value of precision, recall, and F1-score overall, with an Area Under the ROC Curve significantly higher than 0.99. Combined with the fact that the issue of cyberbullying always remains relevant, these results suggest the high feasibility of our suggested approach to the detection of incidents. Given our results, CBNet could be used as a preventative tool for educators, administrators, and community managers to combat cyberbullying behavior and make the online community safer and more welcoming for students. This work suggests the high importance of advanced machine learning approaches to real-world social problems and contributes to the creation of greater digital well-being in university students\u0026rsquo; communities. By employing CBNet, institutions can take proactive measures to mitigate the harmful effects of cyberbullying and cultivate a positive online culture conducive to student success and flourishing.\u003c/p\u003e","manuscriptTitle":"Fostering Supportive Online Communities: Exploring Bystander Intervention in Cyberbullying Prevention","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-20 09:09:04","doi":"10.21203/rs.3.rs-5833561/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-02-17T06:26:17+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-16T07:29:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-15T09:15:48+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-02-08T11:52:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"59192868106473879006403419299407117214","date":"2025-02-02T19:09:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"216365114317611398000573600358827719413","date":"2025-01-30T17:56:33+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"175036358124821973397034928216823187886","date":"2025-01-29T08:43:20+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-01-27T05:30:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"156153792015440130080470887846371753614","date":"2025-01-27T05:22:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"182223509550813417191535079001607346412","date":"2025-01-27T03:39:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"186122382901869091322064288525406670991","date":"2025-01-23T11:12:48+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-01-21T05:07:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-01-21T05:03:08+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-01-21T04:04:50+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-01-17T11:33:21+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-01-15T10:19:39+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f0a642b4-1c53-4340-ba58-e58490816b6b","owner":[],"postedDate":"January 20th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":43020275,"name":"Physical sciences/Mathematics and computing/Computational science"},{"id":43020276,"name":"Physical sciences/Mathematics and computing/Computer science"},{"id":43020277,"name":"Physical sciences/Mathematics and computing/Information technology"},{"id":43020278,"name":"Physical sciences/Mathematics and computing/Scientific data"}],"tags":[],"updatedAt":"2025-07-21T16:09:03+00:00","versionOfRecord":{"articleIdentity":"rs-5833561","link":"https://doi.org/10.1038/s41598-025-09091-y","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-07-15 16:05:36","publishedOnDateReadable":"July 15th, 2025"},"versionCreatedAt":"2025-01-20 09:09:04","video":"","vorDoi":"10.1038/s41598-025-09091-y","vorDoiUrl":"https://doi.org/10.1038/s41598-025-09091-y","workflowStages":[]},"version":"v1","identity":"rs-5833561","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5833561","identity":"rs-5833561","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}