Attention Enhanced BiLSTM for Causal Sentiment Mining in Noisy Social Media Streams

Attention Enhanced BiLSTM for Causal Sentiment Mining in Noisy Social Media Streams | 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 Attention Enhanced BiLSTM for Causal Sentiment Mining in Noisy Social Media Streams MILOUD MIHOUBI, MERIEM ZERKOUK, belkacem Chikhaoui This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7021396/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 31 Jan, 2026 Read the published version in International Journal of Data Science and Analytics → Version 1 posted 11 You are reading this latest preprint version Abstract Social media platforms generate a torrent of short, informal messages rich in slang, emojis, and context dependent expressions, making accurate sentiment classification and real time interpretability a major challenge for organizations. We propose a novel CNN BiLSTM MHSA architecture augmented with a Noise Invariant Contrastive Head (NICH) that simultaneously boosts predictive performance and exposes the lexical triggers of sentiment shifts. Our model combines pretrained GloVe embeddings, a lightweight convolutional frontend for local n-gram detection, a bidirectional LSTM for contextual modelling, and a multihead self attention layer to dynamically highlight sentiment bearing tokens. Training proceeds in two phases: embeddings are first frozen for stable convergence, then fine tuned with AdamW under a cosine decay learning rate schedule. On two benchmarks (1.6 M tweets from Sentiment140 and 50 k IMDB reviews), we outperform classical baselines (LR, SVM, RF) nd exceed deep variants lacking attention or bidirectionality. Ablation confirms MHSA as the primary driver of gains followed by bidirectionality and dropout. Attention reveal causal patterns negators, intensifiers, key emojis that underpin predictions, enabling targeted interventions in marketing, crisis response, and customer service. Our framework thus provides a scalable blueprint for causal text mining in any noisy, high velocity data stream where interpretability and timely decisions are paramount Sentiment Analysis Attention Mechanism Bidirectional LSTM Causal Inference Noisy Web Data Social Media Analytics Noise Invariant Contrastive Head Deep Learning Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 31 Jan, 2026 Read the published version in International Journal of Data Science and Analytics → Version 1 posted Editorial decision: Revision requested 15 Oct, 2025 Reviews received at journal 26 Sep, 2025 Reviewers agreed at journal 25 Aug, 2025 Reviews received at journal 25 Aug, 2025 Reviewers agreed at journal 21 Aug, 2025 Reviewers agreed at journal 19 Aug, 2025 Reviewers agreed at journal 19 Aug, 2025 Reviewers invited by journal 04 Aug, 2025 Editor assigned by journal 19 Jul, 2025 Submission checks completed at journal 02 Jul, 2025 First submitted to journal 01 Jul, 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-7021396","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":495707012,"identity":"e4ee23da-86d7-420f-9c89-f134ba7716f7","order_by":0,"name":"MILOUD MIHOUBI","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4ElEQVRIiWNgGAWjYDACZuYGBgY2IJIAshkMbBgY2AlqYUTRkiYBJAkBqBYGsBaGw4S16LYzNj7mKWOI5pNuPiZdUHC+jp+Zx/AGQ40dTi1mhxmbjXnOMeS2yRxLk55hcFtCspnH2ILhWDI+LW3SvG1ALRI5ZtI8QC0Gh3nMJBgbcDsPXcs5CXuIlnqitRyQMGAGazmM1y+Gc85JALWkJVvPMEiWnHGYrdgi4dhx3FrOHz744E2ZTe78GckHbxf8sePnb2/eeONDTTVOLVAggcZNIKSBkAmjYBSMglEwwgEAo3ZEujUKKrEAAAAASUVORK5CYII=","orcid":"","institution":"Université TÉLUQ","correspondingAuthor":true,"prefix":"","firstName":"MILOUD","middleName":"","lastName":"MIHOUBI","suffix":""},{"id":495707013,"identity":"1269740f-e985-40dd-b1c8-19cbb366c471","order_by":1,"name":"MERIEM ZERKOUK","email":"","orcid":"","institution":"Université TÉLUQ","correspondingAuthor":false,"prefix":"","firstName":"MERIEM","middleName":"","lastName":"ZERKOUK","suffix":""},{"id":495707014,"identity":"deb56d02-194b-424d-9961-2e4f328e2c91","order_by":2,"name":"belkacem Chikhaoui","email":"","orcid":"","institution":"Université TÉLUQ","correspondingAuthor":false,"prefix":"","firstName":"belkacem","middleName":"","lastName":"Chikhaoui","suffix":""}],"badges":[],"createdAt":"2025-07-01 14:38:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7021396/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7021396/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s41060-026-01027-7","type":"published","date":"2026-01-31T15:59:20+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":101690694,"identity":"992f52f3-5d27-4bd0-9a65-65193f08f014","added_by":"auto","created_at":"2026-02-02 16:07:40","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":7673167,"visible":true,"origin":"","legend":"","description":"","filename":"paper.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7021396/v1_covered_00fea69f-df23-4d67-b83c-06dff7a62bfd.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Attention Enhanced BiLSTM for Causal Sentiment Mining in Noisy Social Media Streams","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-data-science-and-analytics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jdsa","sideBox":"Learn more about [International Journal of Data Science and Analytics](http://link.springer.com/journal/41060)","snPcode":"41060","submissionUrl":"https://submission.nature.com/new-submission/41060/3","title":"International Journal of Data Science and Analytics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Sentiment Analysis, Attention Mechanism, Bidirectional LSTM, Causal Inference, Noisy Web Data, Social Media Analytics, Noise Invariant Contrastive Head, Deep Learning","lastPublishedDoi":"10.21203/rs.3.rs-7021396/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7021396/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Social media platforms generate a torrent of short, informal messages rich in slang, emojis, and context dependent expressions, making accurate sentiment classification and real time interpretability a major challenge for organizations. We propose a novel CNN BiLSTM MHSA architecture augmented with a Noise Invariant Contrastive Head (NICH) that simultaneously boosts predictive performance and exposes the lexical triggers of sentiment shifts. Our model combines pretrained GloVe embeddings, a lightweight convolutional frontend for local n-gram detection, a bidirectional LSTM for contextual modelling, and a multihead self attention layer to dynamically highlight sentiment bearing tokens. Training proceeds in two phases: embeddings are first frozen for stable convergence, then fine tuned with AdamW under a cosine decay learning rate schedule. On two benchmarks (1.6 M tweets from Sentiment140 and 50 k IMDB reviews), we outperform classical baselines (LR, SVM, RF) nd exceed deep variants lacking attention or bidirectionality. Ablation confirms MHSA as the primary driver of gains followed by bidirectionality and dropout. Attention reveal causal patterns negators, intensifiers, key emojis that underpin predictions, enabling targeted interventions in marketing, crisis response, and customer service. Our framework thus provides a scalable blueprint for causal text mining in any noisy, high velocity data stream where interpretability and timely decisions are paramount","manuscriptTitle":"Attention Enhanced BiLSTM for Causal Sentiment Mining in Noisy Social Media Streams","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-06 07:13:18","doi":"10.21203/rs.3.rs-7021396/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-16T02:03:50+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-26T20:07:23+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"273690133172136825472949512050129124810","date":"2025-08-25T11:48:25+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-25T06:33:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"105784938693354624211767812748532906347","date":"2025-08-21T09:11:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"297525811663256899267098549465189674533","date":"2025-08-19T12:59:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"119345787310668444000182697355395777817","date":"2025-08-19T10:37:51+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-05T02:06:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-20T01:10:26+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-02T13:08:05+00:00","index":"","fulltext":""},{"type":"submitted","content":"International Journal of Data Science and Analytics","date":"2025-07-01T14:23:49+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-data-science-and-analytics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jdsa","sideBox":"Learn more about [International Journal of Data Science and Analytics](http://link.springer.com/journal/41060)","snPcode":"41060","submissionUrl":"https://submission.nature.com/new-submission/41060/3","title":"International Journal of Data Science and Analytics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"21bbfc5b-8328-45b4-85ea-2db2976d574f","owner":[],"postedDate":"August 6th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-02-02T16:03:47+00:00","versionOfRecord":{"articleIdentity":"rs-7021396","link":"https://doi.org/10.1007/s41060-026-01027-7","journal":{"identity":"international-journal-of-data-science-and-analytics","isVorOnly":false,"title":"International Journal of Data Science and Analytics"},"publishedOn":"2026-01-31 15:59:20","publishedOnDateReadable":"January 31st, 2026"},"versionCreatedAt":"2025-08-06 07:13:18","video":"","vorDoi":"10.1007/s41060-026-01027-7","vorDoiUrl":"https://doi.org/10.1007/s41060-026-01027-7","workflowStages":[]},"version":"v1","identity":"rs-7021396","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7021396","identity":"rs-7021396","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}