Explainable Amharic Emotional Text Classification Using Transfer Learning

Explainable Amharic Emotional Text Classification Using Transfer Learning | 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 Explainable Amharic Emotional Text Classification Using Transfer Learning Demeke Endalie, Yeshimebet Bayu, Tesfa Tegegne This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8392982/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 The widespread use of digital platforms has led to a surge in Amharic emotional comments, highlighting the critical need for robust emotion detection systems in this under-resourced language. To address this, our paper introduces a novel Amharic emotion detection model, fine-tuned from a large language model pre-trained on 17 African languages, including Amharic. Crucially, we integrate Explainable Artificial Intelligence (XAI) into our pipeline to enhance the transparency of the model's decisions. The model was trained, validated, and tested on 14,191, 1,577, and 1,752 expert-labeled Amharic social media comments, categorized into seven emotions: neutral, fear, sadness, joy, anger, surprise, and disgust. Utilizing Local Interpretable Model-agnostic Explanations (LIME) as our XAI framework, the experimental results demonstrate excellent performance, achieving an accuracy of 68.15%, with comparable precision (68.24%), recall (68.15%), and F1-scores (68.17%). The XAI algorithms clearly show how the model arrives at its classifications by calculating probabilities for each category and highlighting relevant textual terms. When compared against several state-of-the-art text classification techniques, including Bidirectional Encoder Representations from Transformers (base variant), Support Vector Machine, Bidirectional Long Short-Term Memory (Bi-LSTM), LSTM, and Convolutional Neural Network (trained from scratch), our fine-tuned model consistently outperforms these approaches. While promising, these results indicate avenues for further improvement. Future research will focus on enhancing performance through training on larger emotional datasets and incorporating code-switching. Amharic NLP Emotion Detection Low-Resource Languages Model Fine-Tuning Social Media Analysis XAI Full Text Additional Declarations No competing interests reported. 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-8392982","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":563307918,"identity":"43bf7a68-30ec-4db6-b981-e0bfe4211585","order_by":0,"name":"Demeke Endalie","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFElEQVRIiWNgGAWjYDCCAwwMjA1gVgLjgQQGmwSQCAMPkVoYgFrSEFpwakPRwsBwmLAWvuPdiQ9nVDDImbMnPzjwoOJ8Ht+NBMYHb9sY5OxxaJE8c3az4YYzDMaWPc8MDiScuV0seSOB2XBuG4MxLlsMbuRuk3zYxpC44UaCwYHEttsgBps0L1CkB5eW+2+3/3z4j6F+w430DwcS/50DaWH/DdRSj1PLDd5tjBsbGBIMbuQAbWk4ALaFGaglAZfDJM/kbpaccUwC6J83BQcSjiUXS5552Cw555yEYc8BXCF2duPHnhobeYPj6Rsf/qixy+M7nnzww5syG3n2BhzWQIAEMgccTxLYFY6CUTAKRsEoIAoAALpHbBXZSucpAAAAAElFTkSuQmCC","orcid":"","institution":"Faculty of Computing, Bahir Dar Institute of Technology, Bahir Dar University","correspondingAuthor":true,"prefix":"","firstName":"Demeke","middleName":"","lastName":"Endalie","suffix":""},{"id":563307920,"identity":"1cc0511e-bb8f-4b52-be99-0aca61229ae1","order_by":1,"name":"Yeshimebet Bayu","email":"","orcid":"","institution":"Department of Computer Science, Debre Markos Institute of Technology, Debre Markos University","correspondingAuthor":false,"prefix":"","firstName":"Yeshimebet","middleName":"","lastName":"Bayu","suffix":""},{"id":563307925,"identity":"a21f52a6-c3c0-4993-83e3-85aa94c9d4cb","order_by":2,"name":"Tesfa Tegegne","email":"","orcid":"","institution":"Faculty of Computing, Bahir Dar Institute of Technology, Bahir Dar University","correspondingAuthor":false,"prefix":"","firstName":"Tesfa","middleName":"","lastName":"Tegegne","suffix":""}],"badges":[],"createdAt":"2025-12-18 08:38:32","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8392982/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8392982/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100121422,"identity":"b9f99866-d788-422d-9a0b-febb3278e6d7","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":323561,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/ae3a33f81e04465a3cc0c401.docx"},{"id":100365384,"identity":"c7fc9099-199d-49c5-8a77-432dc78e1ae4","added_by":"auto","created_at":"2026-01-16 07:55:10","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5167,"visible":true,"origin":"","legend":"","description":"","filename":"0c89c378187b4e299b008187da84f45a.json","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/8fd626696cb5c0c8523f0724.json"},{"id":100365653,"identity":"e46fde00-acc8-436c-83c0-e82d37a6d24c","added_by":"auto","created_at":"2026-01-16 07:55:29","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":112502,"visible":true,"origin":"","legend":"","description":"","filename":"0c89c378187b4e299b008187da84f45a1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/a37b57b07e2e175cff2b43ca.xml"},{"id":100121426,"identity":"3c59d2f5-46db-449a-9afe-0aac411caaac","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":99241,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/d2a09549ffdf4a5506305f9a.png"},{"id":100121418,"identity":"8ea9423a-c916-4ab6-859f-80ca50762741","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":18876,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/0aa5ef443c13198745a3ce46.png"},{"id":100121417,"identity":"268bec23-762d-4619-8e56-a51250662de9","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9510,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/29c3a2a16d6ae55329aebea3.png"},{"id":100121416,"identity":"1cab2a31-298b-476a-9976-cd5c3d2348d2","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4634,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/e72574a4d047bcf3911b0ad6.png"},{"id":100365699,"identity":"506e0746-3d00-48b3-87b3-fff96d7f09f0","added_by":"auto","created_at":"2026-01-16 07:55:32","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5829,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/3586a1d9d6e6f916d6a9a6cb.png"},{"id":100365292,"identity":"12151b81-32e6-404e-b7ee-86207ea3d379","added_by":"auto","created_at":"2026-01-16 07:55:00","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5922,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/a871accede95ceaec739fa3e.png"},{"id":100121420,"identity":"fce15390-0dac-4b85-9914-c291cce0f3a5","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6615,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/a3da53dfde0d7ba84a04bb8b.png"},{"id":100121425,"identity":"e76cd7e4-c5a2-4c78-85ed-58cf61a3ebc4","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":54608,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/00a51560e6b6be73dca28f16.png"},{"id":100365440,"identity":"a68732a2-94e6-488e-b411-7ba4b6f3bb6f","added_by":"auto","created_at":"2026-01-16 07:55:13","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53277,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/c71abb6f11980e45503f8edb.png"},{"id":100367595,"identity":"0935c4a2-8c6e-4053-ac3c-ab040a99b673","added_by":"auto","created_at":"2026-01-16 07:57:08","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":41803,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/f7f41c34492c62550d158720.png"},{"id":100365682,"identity":"57592fc8-8395-407f-b5e6-e94f33af27de","added_by":"auto","created_at":"2026-01-16 07:55:30","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6680,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/47fb395666cebe249e084b5e.png"},{"id":100366039,"identity":"ca6909c2-76bf-49fe-8360-49f386ba711e","added_by":"auto","created_at":"2026-01-16 07:55:53","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3211,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/90ad61e02fa021eb303cc644.png"},{"id":100121431,"identity":"9024ba4b-aedf-4489-8bd8-ae9665d1458f","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2498,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/c75f97d58d70da82dc379ab9.png"},{"id":100367667,"identity":"782a7182-6330-45e4-8893-8cf6c7bf0994","added_by":"auto","created_at":"2026-01-16 07:57:13","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2366,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/7b86a9f2342203a76945d053.png"},{"id":100121433,"identity":"05b8d866-3d8b-4bbf-8988-ed908d70acc6","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2828,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/6b25a423f16064875c2935d6.png"},{"id":100121428,"identity":"83f020b8-e6d3-48dc-aa25-3c67aea766e0","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2977,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/58182eca77d3e38e8c053088.png"},{"id":100121436,"identity":"2491d5e9-b3f4-4886-8aad-07bf18d905c0","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17658,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/68a0be454c27952f2c2aa799.png"},{"id":100365934,"identity":"7747cabd-f1f6-480a-be95-9e40f47d82ad","added_by":"auto","created_at":"2026-01-16 07:55:46","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17473,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/f48ab1563b1d83b686712976.png"},{"id":100121438,"identity":"a60e243a-3649-47b3-b6f0-c4804f574ecb","added_by":"auto","created_at":"2026-01-13 08:46:38","extension":"xml","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":113000,"visible":true,"origin":"","legend":"","description":"","filename":"0c89c378187b4e299b008187da84f45a1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/4a883df122055ceb0d311d96.xml"},{"id":100121437,"identity":"f44ad076-fe41-4e8f-889a-168427fee0cb","added_by":"auto","created_at":"2026-01-13 08:46:37","extension":"html","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121857,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1/06c517174b4c966cf6160443.html"},{"id":100382258,"identity":"a41ad00e-12fc-4bc7-83d4-aa49c3f977d1","added_by":"auto","created_at":"2026-01-16 10:41:51","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":703585,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8392982/v1_covered_ec2373cc-4811-4a9d-a4a1-163c676d6192.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Explainable Amharic Emotional Text Classification Using Transfer Learning","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":"Amharic NLP, Emotion Detection, Low-Resource Languages, Model Fine-Tuning, Social Media Analysis, XAI","lastPublishedDoi":"10.21203/rs.3.rs-8392982/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8392982/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe widespread use of digital platforms has led to a surge in Amharic emotional comments, highlighting the critical need for robust emotion detection systems in this under-resourced language. To address this, our paper introduces a novel Amharic emotion detection model, fine-tuned from a large language model pre-trained on 17 African languages, including Amharic. Crucially, we integrate Explainable Artificial Intelligence (XAI) into our pipeline to enhance the transparency of the model's decisions. The model was trained, validated, and tested on 14,191, 1,577, and 1,752 expert-labeled Amharic social media comments, categorized into seven emotions: neutral, fear, sadness, joy, anger, surprise, and disgust. Utilizing Local Interpretable Model-agnostic Explanations (LIME) as our XAI framework, the experimental results demonstrate excellent performance, achieving an accuracy of 68.15%, with comparable precision (68.24%), recall (68.15%), and F1-scores (68.17%). The XAI algorithms clearly show how the model arrives at its classifications by calculating probabilities for each category and highlighting relevant textual terms. When compared against several state-of-the-art text classification techniques, including Bidirectional Encoder Representations from Transformers (base variant), Support Vector Machine, Bidirectional Long Short-Term Memory (Bi-LSTM), LSTM, and Convolutional Neural Network (trained from scratch), our fine-tuned model consistently outperforms these approaches. While promising, these results indicate avenues for further improvement. Future research will focus on enhancing performance through training on larger emotional datasets and incorporating code-switching.\u003c/p\u003e","manuscriptTitle":"Explainable Amharic Emotional Text Classification Using Transfer Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-13 08:46:26","doi":"10.21203/rs.3.rs-8392982/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":"e77be9c9-ee44-43a6-a621-6af5c91314d5","owner":[],"postedDate":"January 13th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-13T08:46:26+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-13 08:46:26","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8392982","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8392982","identity":"rs-8392982","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}