A Novel Hybrid Deep Learning Model for Aspect-Based Sentiment Analysis (ABSA)

A Novel Hybrid Deep Learning Model for Aspect-Based Sentiment Analysis (ABSA) | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article A Novel Hybrid Deep Learning Model for Aspect-Based Sentiment Analysis (ABSA) PRIYANKA SHAKYA, AANCHAL SINGH, Ashish Kumar Mishra This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9354202/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract A smooth sentiment analysis task, called Aspect-Based Sentiment Analysis (ABSA), focuses on pinpointing specific aspect phrases in a review and determining the sentiment polarity associated with each one. Unlike traditional sentiment analysis, which only captures the overall opinion expressed in a text, ABSA aims to extract detailed insights by evaluating sentiments tied to distinct components of a product, service, or topic. ABSA involves gathering and analysing public opinions, feelings, and attitudes, providing valuable information for businesses and organizations looking to improve customer satisfaction and decision-making processes. One major difficulty lies in the presence of irrelevant or noisy information in the review text, which can obscure the true sentiment. Additionally, accurately detecting the boundaries of aspect terms is complex, as these terms often appear in various forms and contexts. Another challenge is the subtle and implicit nature of sentiment expression, which can be further complicated by the use of negations, sarcasm, and rhetorical devices that alter the intended sentiment. The suggested solution uses a hybrid model that gives sentiment polarity to both the individual aspect phrases and the sentence as a whole to overcome these difficulties. It combines two cutting-edge neural architectures: a Long Short-Term Memory (LSTM) and a Graph Attention Network (GAT). LSTM networks are well-suited for determining the overall sentiment of a sentence due to their strength in modelling the sequential structure of textual data. In the meantime, the GAT is quite good at simulating the connections between words in a sentence, which enables it to draw attention to the pertinent context of each aspect term. The proposed system achieves an accuracy of 94.22% on the laptop dataset, and its results are compared with baselines as well as existing works on ABSA. The comparison shows that the performance of the proposed approach is better than the existing approaches in the literature. Artificial Intelligence and Machine Learning GAT ABSA LSTM. Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9354202","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":619529350,"identity":"daf07042-3de6-44d0-86c9-4e7e86e904c4","order_by":0,"name":"PRIYANKA SHAKYA","email":"","orcid":"","institution":"Rajkiya Engineering College, Ambedkar Nagar, Uttar Pradesh, India","correspondingAuthor":false,"prefix":"","firstName":"PRIYANKA","middleName":"","lastName":"SHAKYA","suffix":""},{"id":619529351,"identity":"3c375c7a-1786-489d-af67-ebe23bf611e5","order_by":1,"name":"AANCHAL SINGH","email":"","orcid":"","institution":"Rajkiya Engineering College, Ambedkar Nagar, Uttar Pradesh, India","correspondingAuthor":false,"prefix":"","firstName":"AANCHAL","middleName":"","lastName":"SINGH","suffix":""},{"id":619529353,"identity":"22acf63c-b01f-4e9f-86d3-9b9d0c3cfb38","order_by":2,"name":"Ashish Kumar Mishra","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8ElEQVRIie2RsYoCMRCGJwS02XXb2OgrjFwlKPcqHgdXrfoKgpBrPG218i3EMmHaPWwX0twj7HauWJiglu7udYL5mv8v5iPMBMDjeUoCUGzmSvSnMhuNZn0lYHrtFF5DgZvCKXBZpUSrX62LPUxtUTQ47botDizL48eKSCcjChPor10ZL01PcuDtza7kmTRAYhIQBSCNF4ZZpcHDEqV7SFAXTokUUn9h3isVVDGq0CkQI8HRfFQqvdROhlKgSL9G+mdmPiVn89JdOofkLS/kAKMVUXY8m+H2e66zvGz9K+Ia9g5w/6a6nP8z7PF4PK/CBQmeVDG6RtcIAAAAAElFTkSuQmCC","orcid":"","institution":"Rajkiya Engineering College, Ambedkar Nagar, Uttar Pradesh, India","correspondingAuthor":true,"prefix":"","firstName":"Ashish","middleName":"Kumar","lastName":"Mishra","suffix":""}],"badges":[],"createdAt":"2026-04-08 09:04:27","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9354202/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9354202/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106724579,"identity":"e90402b8-4ab5-40ee-ac0c-4cb4331a9f97","added_by":"auto","created_at":"2026-04-12 18:28:41","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":736790,"visible":true,"origin":"","legend":"","description":"","filename":"Updatedpaper1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9354202/v1_covered_34d3d3e2-4c17-4502-bebd-b373a36381f3.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eA Novel Hybrid Deep Learning Model for Aspect-Based Sentiment Analysis (ABSA)\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"GAT, ABSA, LSTM.","lastPublishedDoi":"10.21203/rs.3.rs-9354202/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9354202/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eA smooth sentiment analysis task, called Aspect-Based Sentiment Analysis (ABSA), focuses on pinpointing specific aspect phrases in a review and determining the sentiment polarity associated with each one. Unlike traditional sentiment analysis, which only captures the overall opinion expressed in a text, ABSA aims to extract detailed insights by evaluating sentiments tied to distinct components of a product, service, or topic. ABSA involves gathering and analysing public opinions, feelings, and attitudes, providing valuable information for businesses and organizations looking to improve customer satisfaction and decision-making processes. One major difficulty lies in the presence of irrelevant or noisy information in the review text, which can obscure the true sentiment. Additionally, accurately detecting the boundaries of aspect terms is complex, as these terms often appear in various forms and contexts. Another challenge is the subtle and implicit nature of sentiment expression, which can be further complicated by the use of negations, sarcasm, and rhetorical devices that alter the intended sentiment. The suggested solution uses a hybrid model that gives sentiment polarity to both the individual aspect phrases and the sentence as a whole to overcome these difficulties. It combines two cutting-edge neural architectures: a Long Short-Term Memory (LSTM) and a Graph Attention Network (GAT). LSTM networks are well-suited for determining the overall sentiment of a sentence due to their strength in modelling the sequential structure of textual data. In the meantime, the GAT is quite good at simulating the connections between words in a sentence, which enables it to draw attention to the pertinent context of each aspect term. The proposed system achieves an accuracy of 94.22% on the laptop dataset, and its results are compared with baselines as well as existing works on ABSA. The comparison shows that the performance of the proposed approach is better than the existing approaches in the literature.\u003c/p\u003e","manuscriptTitle":"A Novel Hybrid Deep Learning Model for Aspect-Based Sentiment Analysis (ABSA)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-09 03:58:07","doi":"10.21203/rs.3.rs-9354202/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"369043c5-ef6c-4aca-961e-f24ee5732ba8","owner":[],"postedDate":"April 9th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":65918235,"name":"Artificial Intelligence and Machine Learning"}],"tags":[],"updatedAt":"2026-04-09T03:58:07+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-09 03:58:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9354202","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9354202","identity":"rs-9354202","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}