Elamur Research⌕
☰
home / papers / 2026 / TamilPoliSent 2026: A Shared Task report on Multiclass Political Sentiment Analysis in Tamil

TamilPoliSent 2026: A Shared Task report on Multiclass Political Sentiment Analysis in Tamil

Paper · 2026 · Mani Vegupatti, Kishore Kumar Ponnusamy, Bharathi Raja Chakravarthi et al.

Summary

TamilPoliSent 2026, held at DravidianLangTech@ACL 2026, is the second edition of a shared task on fine-grained political sentiment classification in Tamil. Tamil is a morphologically rich, low-resource language, and its social media text is noisy, frequently code-mixed and often sarcastic. The organizers gathered Tamil posts from X using the twscrape tool, searching hashtags tied to Tamil Nadu parties and elections. Every post was given exactly one tag from a set of seven: Neutral, Positive, Sarcastic, Negative, Substantiated, Opinionated, or None of the above. Each post was labeled by two annotators, and only posts with matching labels were kept. This produced 5,440 items, split into training, development and an unlabeled test portion. Twenty-two teams submitted systems, including TF-IDF classifiers, fine-tuned Indic and multilingual transformers, lexical–contextual hybrids, ensembles and imbalance-aware training. Systems were ranked by macro-F1, and the top entry reached only 0.3935. This low ceiling shows that skewed labels, sarcasm and overlapping categories remain difficult. The task provides a public benchmark and a snapshot of current methods for analyzing Tamil political discourse.

Key points

  • The dataset contains 5,440 Tamil tweets: 4,352 for training, 544 for development and 544 for testing. Tweets were collected via hashtags such as #tnelection2024, #dmk, #admk and #ntk, and the task is hosted on Codabench.
  • The classes are strongly imbalanced. Opinionated is the largest class (1,685 tweets) and Sarcastic is second (1,011), while None of the above is the smallest (216).
  • Nine annotators took part. Each tweet received two independent labels, conflicting cases were dropped, and agreement was measured with Krippendorff's alpha.
  • Of the 22 teams, the best scored macro-F1 0.3935. Several other systems scored between 0.35 and 0.37.
  • The strongest results came from transformers combined with class-balanced objectives and hybrid lexical–contextual features. The authors point to data augmentation, culturally informed representations and explainability as future directions.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Ollscoil na Gaillimhe – University of GalwayGalway · Ireland25 works · 2,751 citations🎓Central University of Tamil NaduThiruvarur · India3 works · 15 citations🎓Kerala University of Digital Sciences, Innovation and TechnologyThiruvananthapuram · India2 works · 4 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 works

🧭 Topics 1Research subjects and languages the work is about.Sentiment & emotion · 1,035 works

Cite

Cite as Mani Vegupatti, Kishore Kumar Ponnusamy, Bharathi Raja Chakravarthi, Saranya Rajiakodi, Thenmozhi Durairaj, Prasanna Kumar Kumaresan and Sathiyaraj Thangasamy. 2026. TamilPoliSent 2026: A Shared Task report on Multiclass Political Sentiment Analysis in Tamil. Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.dravidianlangtech-1.15

@inproceedings{vegupatti2026tamilpolisent,
  title = {{TamilPoliSent 2026: A Shared Task report on Multiclass Political Sentiment Analysis in Tamil}},
  author = {Mani Vegupatti, Kishore Kumar Ponnusamy, Bharathi Raja Chakravarthi, Saranya Rajiakodi, Thenmozhi Durairaj, Prasanna Kumar Kumaresan and Sathiyaraj Thangasamy},
  booktitle = {Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages},
  year = {2026},
  doi = {10.18653/v1/2026.dravidianlangtech-1.15},
  url = {https://doi.org/10.18653/v1/2026.dravidianlangtech-1.15}
}

References

31 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ramaswami Abishek, Rohith P P, Kogilavani Shanmugavadivel, Gothai E, and Malliga Subramanian. 2026. Dravidian Decoders@DravidianLangTech 2026: Political Multiclass Sentiment Analysis of Tamil X (Twitter) via Asymmetric Weighted Cross-Entropy. In Proceedings of the Sixth Workshop on Speech, Vision, a… find
  2. Tamara Amjad Al-Qablan, Mohd Halim Mohd Noor, M. Al-Betar, and A. Khader. 2023. A survey on sentiment analysis and its applications. Neural Computing and Applications, 35:21567–21601. find
  3. S. Anbukkarasi and S. Varadhaganapathy. 2020. Analyzing sentiment in tamil tweets using deep neural network. ICCMC, pages 449–453. find
  4. Karunanidhi Arunaggiri Pandian and Arumugam Prabalakshmi. 2026. CHMOD_777@DravidianLangTech 2026: LLM-Augmented Transformer Fine-Tuning for Tamil Political Sentiment Analysis. In Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages. Association for C… find
  5. Akshat Bakliwal, Jennifer Foster, Jennifer van der Puil, Ron O’Brien, Lamia Tounsi, and Mark Hughes. 2013. Sentiment analysis of political tweets: Towards an accurate classifier. In Proceedings of the Workshop on Language Analysis in Social Media, pages 49– 58, Atlanta, Georgia. Association for Comp… find
  6. Adam Bermingham and Alan Smeaton. 2011. On using Twitter to monitor political sentiment and predict election results. In Proceedings of the Workshop on Sentiment Analysis where AI meets Psychology (SAAIP 2011), pages 2–10, Chiang Mai, Thailand. Asian Federation of Natural Language Processing. find
  7. Bharathi Raja Chakravarthi, Saranya Rajiakodi, Elizabeth Sherly, Thenmozhi Durairaj, Sathiyaraj Thangasamy, Ratnasingam Sakuntharaj, Prasanna Kumar Kumaresan, Kishore Kumar Ponnusamy, Arunaggiri Pandian Karunanidhi, and Rohan R. 2025. Overview on political multiclass sentiment analysis of tamil x (t… find
  8. Jessica Chung and Eni Mustafaraj. 2011. Can collective sentiment expressed on twitter predict political elections? In Proceedings of the AAAI Conference on Artificial Intelligence, volume 25, pages 1770–1771. find
  9. Alfred et al. Hermida. 2012. Share, like, recommend: Decoding the social media news consumer. Journalism Studies. find
  10. CS Kanimozhi Selvi, NS Inigashree, AG Moneissh, and J Kavinraj. 2026. TamilEcho@DravidianLangTech 2026: Hybrid XLM-RoBERTa with Sarcasm-Aware Feature Fusion for Political Multiclass Sentiment Analysis in Tamil X. In Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for D… find
  11. Sainik Kumar Mahata, Dipankar Das, and Sivaji Bandyopadhyay. 2020. Sentiment classification of code-mixed tamil-english tweets using birnn. arXiv preprint arXiv:2010.10111. — [arXiv:2010.10111](https://arxiv.org/abs/2010.10111) find
  12. S. Manoj. 2024. A survey on sentiment analysis in tamil: Critical analysis. In Annals of Computer Science and Information Systems, volume 42, pages 49–62. find
  13. TV Mitharshana, S Shanthi, V Lavana, and R Kaviya Varma. 2026. Byte-Builders@DravidianLangTech 2026: Transformer-Based Weighted Ensemble for Political Multiclass Sentiment Analysis of Tamil X (Twitter) Comments. In Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for Dr… find
  14. Marouane Birjali, Mohammed Kasri, and A. B. Hssane. 2021. A comprehensive survey on sentiment analysis: Approaches, challenges and trends. Knowl. Based Syst., 226:107134. find
  15. Kandasamy Naren Karthik, Aakash Balasubramanian, Hemnath Dillibabu, Aadhithya Srinivasan Rajashree, and Vijaya Lakshmi A. 2026. Sen-Tamizh@DravidianLangTech 2026: A Specialist-Augmented Transformer Ensemble for Multiclass Political Sentiment Analysis of Tamil X Comments. In Proceedings of the Sixth… find
  16. Bharathi Raja Chakravarthi, R. Priyadharshini, Vigneshwaran Muralidaran, Shardul Suryawanshi, Navya Jose, E. Sherly, and John P. McCrae. 2020. Overview of the track on sentiment analysis for dravidian languages in code-mixed text. Proceedings of the 12th Annual Meeting of the Forum for Information R… find
  17. Dinh Khac Phuc Nguyen and Dang Van Thin. 2026. PhucNguyen@DravidianLangTech 2026: Political Multiclass Sentiment Analysis with XLM-RoBERTa and Low-Rank Adaptation. In Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages. Association for Computational…
  18. Ruba Priyadharshini, Bharathi Raja Chakravarthi, Sajeetha Thavareesan, Dhivya Chinnappa, Durairaj Thenmozhi, and Rahul Ponnusamy. 2021. Overview of the dravidiancodemix 2021 shared task on sentiment detection in tamil, malayalam, and kannada. In Proceedings of the 13th Annual Meeting of the Forum fo… find
  19. M. Wankhade, A. C. S. Rao, and C. Kulkarni. 2022. A survey on sentiment analysis methods, applications, and challenges. Artificial Intelligence Review, 55(8):5731–5780. find
  20. Orsolya Ring, Martina Katalin Szabó, Csenge Guba, Bendegúz Váradi, and István Üveges. 2024. Approaches to sentiment analysis of hungarian political news at the sentence level. Language Resources and Evaluation, 58(4):1233–1261. find
  21. Kogilavani et al. Shanmugavadivel. 2022. An analysis of machine learning models for sentiment analysis of tamil code-mixed data. Computer Speech & Language, 76:101407. find
  22. Murugan Shanthi, Anbuaruvi R, and Anuradha C. 2026. TrailBlazer@DravidianLangTech 2026: A Comparative Study of TF-IDF SVM and XLM-RoBERTa for Political Multiclass Text Classification. In Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages. Associati… find
  23. A. Sharmista and M. Ramaswami. 2020. Sentiment analysis on tamil reviews as products in social media using machine learning techniques: A novel study. Madurai Kamaraj University, Madurai-625 021. Technical Report. find
  24. Kian Long Tan, C. Lee, and K. Lim. 2023. A survey of sentiment analysis: Approaches, datasets, and future research. Applied Sciences. find
  25. Sajeetha Thavareesan and Sinnathamby Mahesan. 2019. Sentiment analysis in tamil texts: A study on machine learning techniques. In Proceedings of ICIIS, pages 320–325. find
  26. Sho Tsugawa and Hiroyuki Ohsaki. 2017. On the relation between message sentiment and its virality on social media. Social network analysis and mining, 7:1–14. find
  27. Andranik et al. Tumasjan. 2010. Predicting elections with twitter. In AAAI ICWSM, pages 178–185. find
  28. István Üveges and Orsolya Ring. 2023. Hunembert: a fine-tuned bert-model for classifying sentiment and emotion in political communication. IEEE Access, 11:60267–60278. find
  29. Mani Vegupatti, Kishore Kumar Ponnusamy, Bharathi Raja Chakravarthi, Saranya Rajiakodi, Durairaj Thenmozhi, Prasanna Kumar Kumaresan, and Sathiyaraj Thangasamy. 2026. find
  30. TamilPoliSent 2026: A Shared Task report on Multiclass Political Sentiment Analysis in Tamil. In Proceedings of the Sixth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages. Association for Computational Linguistics.
  31. Alcides Velasquez. 2012. Social media and online political discussion. New Media & Society. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/vegupatti-2026-tamilpolisent-2026-shared-task-report/