Elamur Research⌕
☰
home / papers / 2013 / Identifying Opinion Subgroups in Arabic Online Discussions

Identifying Opinion Subgroups in Arabic Online Discussions

Paper · 2013 · Amjad Abu-Jbara, Ben King, Mona Diab et al.

Summary

The paper addresses how participants in an Arabic online debate split into camps with opposing views. The system takes a forum thread and outputs a subgroup label for every discussant. The pipeline has five components:

  • converting text to Buckwalter encoding;
  • tokenization and NP detection with AMIRAN;
  • word-level polarity from the Sifat lexicon, with context-sensitive sentence polarity from SAMAR;
  • picking opinion targets among noun phrases;
  • adding LDA topic profiles as a latent similarity signal.

Discussants are then represented in one of two ways. In the first, each is a vector of per-target mention and polarity counts plus topic features, and these vectors are clustered. In the second, each is a node in a signed agreement graph, which is partitioned. Evaluation uses 36 mostly-MSA debates from the Naqeshny site, where each user picks a side, so labels come for free. Clustering the vectors with EM worked best and beat graph partitioning. Removing either feature family hurt performance. The work is an early demonstration that opinion-based community detection can be done for Arabic with available resources.

Key points

  • Data: 36 debates from Naqeshny with 711 posts by 326 users, averaging about 19.75 posts and 13.08 participants per debate. Debates were filtered to mostly MSA because the tools were trained on MSA.
  • Resources used:
    • Sifat, a lexicon of 3982 Arabic adjectives tagged positive, negative or neutral.
    • SAMAR, a subjectivity/sentiment classifier. Its reported accuracy is 84–95% for subjectivity and 65–81% for polarity.
    • AMIRAN, used for tokenization and NP chunking.
  • Discussant vector: three counts per target (total, positive, negative) plus a 100-topic LDA distribution, giving 3·t + 100 dimensions.
  • Signed network:
    • An edge links two users who share at least one target. Its sign reflects whether they agree on more shared targets than they disagree on, using each user's majority polarity per target.
    • Partitioning uses the Dorian-Mrvar (1996) objective with α = 0.5 and 3 random restarts of greedy node moves.
  • Results:
    • EM clustering was best, at roughly 0.77 Purity, 0.76 F-measure and 0.50 Entropy.
    • It outperformed K-means, Farthest First, signed-network partitioning, and both single-feature ablations.
    • Differences were significant at 0.05 (2-tailed paired t-test).

Interesting ideas

  • I1. Debate platforms where users explicitly choose a side yield gold group labels with no annotation cost, which makes evaluation of unsupervised grouping cheap.
  • I2. Only noun phrases mentioned by at least two different participants count as opinion targets, which removes incidental phrases.
  • I3. Each opinion is attached to the nearest NP in the sentence, a crude linear-distance heuristic.
  • I4. Agreement between people is derived indirectly by comparing their majority polarity on shared targets, then turned into a signed graph and partitioned by minimizing intra-group negative and inter-group positive edges.
  • I5. Ablating explicit opinion features against latent topic features showed that each contributes something the other lacks.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (learn): Look for similar self-labeled sources to serve as free evaluation sets for stance or role assignment, for example texts where speakers declare their position.
  • I2 (world): A cross-speaker recurrence filter could help pick salient entities in multi-character texts before building the world model.
  • I3 (frames): Keep this nearest-NP rule as a negative-control baseline, so we can show that tree-based attachment of evaluation to its target in prag actually improves on it.
  • I4 (world): The same signed-graph partitioning could split characters in fables or tales into allies and opponents, based on their stated attitudes toward shared entities.
  • I5 (learn): When adding any implicit signal next to explicit frame-based features, run paired ablations with significance tests to prove it adds value and is not just noise.

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

🎓University of MichiganAnn Arbor · United States70 works · 13,693 citations🎓George Washington UniversityWashington · United States51 works · 8,148 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksPhysics and Astronomy · 44 worksStatistical and Nonlinear Physics · 40 works

🧭 Topics 1Research subjects and languages the work is about.Semitic languages · 2,908 works

Cite

Cite as Amjad Abu-Jbara, Ben King, Mona Diab and Dragomir Radev. 2013. Identifying Opinion Subgroups in Arabic Online Discussions. Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers).

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{abujbara2013identifying,
  title = {{Identifying Opinion Subgroups in Arabic Online Discussions}},
  author = {Amjad Abu-Jbara, Ben King, Mona Diab and Dragomir Radev},
  booktitle = {Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)},
  year = {2013},
  url = {https://aclanthology.org/P13-2144/}
}

References

42 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ahmed Abbasi, Hsinchun Chen, and Arab Salem. 2008. Sentiment analysis in multiple languages: Feature selection for opinion classification in web forums. ACM Trans. Inf. Syst., 26(3):12:1–12:34, June. find
  2. Muhammad Abdul-Mageed and Mona Diab. 2011. Subjectivity and sentiment annotation of modern standard arabic newswire. In Proceedings of the 5th Linguistic Annotation Workshop, pages 110– 118, Portland, Oregon, USA, June. Association for Computational Linguistics.
  3. Muhammad Abdul-Mageed and Mona Diab. 2012a. Awatif: A multi-genre corpus for modern standard arabic subjectivity and sentiment analysis. In Nicoletta Calzolari (Conference Chair), Khalid Choukri, Thierry Declerck, Mehmet Ugur Dogan, Bente Maegaard, Joseph Mariani, Jan Odijk, and Stelios Piperidis, editors, Proceedings of the Eight International Conference on Language Resources and Evaluation (LREC’12), Istanbul, Turkey, may. European Language Resources Association (ELRA).
  4. Muhammad Abdul-Mageed and Mona Diab. 2012b. Toward building a large-scale arabic sentiment lexicon. In Proceedings of the 6th International Global Word-Net Conference, Matsue, Japan. find
  5. Muhammad Abdul-Mageed, Sandra Kübler, and Mona Diab. 2012a. Samar: a system for subjectivity and sentiment analysis of arabic social media. In Proceedings of the 3rd Workshop in Computational Approaches to Subjectivity and Sentiment Analysis, WASSA ’12, pages 19–28, Stroudsburg, PA, USA. Association for Computational Linguistics.
  6. Muhammad Abdul-Mageed, Sandra Kuebler, and Mona Diab. 2012b. Samar: A system for subjectivity and sentiment analysis of arabic social media. In Proceedings of the 3rd Workshop in Computational Approaches to Subjectivity and Sentiment Analysis, pages 19–28, Jeju, Korea, July. Association for Computational Linguistics.
  7. Amjad Abu-Jbara and Dragomir Radev. 2012. Subgroup detection in ideological discussions. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, Jeju, Korea, July. The Association for Computational Linguistics. ↗
  8. Pranav Anand, Marilyn Walker, Rob Abbott, Jean E. Fox Tree, Robeson Bowmani, and Michael Minor. 2011. Cats rule and dogs drool!: Classifying stance in online debate. In Proceedings of the 2nd Workshop on Computational Approaches to Subjectivity and Sentiment Analysis (WASSA 2.011), pages 1–9, Portland, Oregon, June. Association for Computational Linguistics.
  9. Carmen Banea, Rada Mihalcea, and Janyce Wiebe. 2008. A bootstrapping method for building subjectivity lexicons for languages with scarce resources. In LREC’08. find
  10. Steven Bethard, Hong Yu, Ashley Thornton, Vasileios Hatzivassiloglou, and Dan Jurafsky. 2004. Automatic extraction of opinion propositions and their holders. In 2004 AAAI Spring Symposium on Exploring Attitude and Affect in Text, page 2224. find
  11. Tim Buckwalter. 2004. Issues in arabic orthography and morphology analysis. In Proceedings of the Workshop on Computational Approaches to Arabic Script-based Languages, Semitic ’04, pages 31–34, Stroudsburg, PA, USA. Association for Computational Linguistics.
  12. Sanjoy Dasgupta. 2002. Performance guarantees for hierarchical clustering. In 15th Annual Conference on Computational Learning Theory, pages 351–363. Springer. find
  13. Pradeep Dasigi, Weiwei Guo, and Mona Diab. 2012. Genre independent subgroup detection in online discussion threads: A study of implicit attitude using textual latent semantics. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 65–69, Jeju Island, Korea, July. Association for Computational Linguistics. find
  14. A. P. Dempster, N. M. Laird, and D. B. Rubin. 1977. Maximum likelihood from incomplete data via the em algorithm. JOURNAL OF THE ROYAL STATIS-TICAL SOCIETY, SERIES B, 39(1):1–38. ↗
  15. Mona Diab. 2009. Second generation tools (amira 2.0): Fast and robust tokenization, pos tagging, and base phrase chunking. In Khalid Choukri and Bente Maegaard, editors, Proceedings of the Second International Conference on Arabic Language Resources and Tools, Cairo, Egypt, April. The MEDAR Consortium. find
  16. Patrick Doreian and Andrej Mrvar. 1996. A partitioning approach to structural balance. Social Networks, 18(2):149–168. ↗
  17. Gregory Grefenstette, Yan Qu, James G Shanahan, and David A Evans. 2004. Coupling niche browsers and affect analysis for an opinion mining application. In Proceedings of RIAO, volume 4, pages 186– 194. Citeseer. find
  18. Weiwei Guo and Mona Diab. 2012. Modeling sentences in the latent space. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 864–872, Jeju Island, Korea, July. Association for Computational Linguistics. ↗
  19. Ahmed Hassan and Dragomir Radev. 2010. Identifying text polarity using random walks. In ACL’10. find
  20. Ahmed Hassan, Vahed Qazvinian, and Dragomir Radev. 2010. What’s with the attitude?: identifying sentences with attitude in online discussions. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, pages 1245–1255. find
  21. Ahmed Hassan, Amjad Abu-Jbara, Rahul Jha, and Dragomir Radev. 2011. Identifying the semantic orientation of foreign words. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies: short papers - Volume 2, HLT ’11, pages 592– 597, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  22. Ahmed Hassan, Amjad Abu-Jbara, and Dragomir Radev. 2012a. Extracting signed social networks from text. In Workshop Proceedings of TextGraphs-7: Graph-based Methods for Natural Language Processing, pages 6–14, Jeju, Republic of Korea, July. Association for Computational Linguistics. find
  23. Ahmed Hassan, Amjad Abu-Jbara, and Dragomir Radev. 2012b. Signed attitude networks: Predicting positive and negative links using linguistic analysis. In Submitted to the Conference on Emprical Methods in Natural Language Processing, Jeju, Korea, July. The Association for Computational Linguistics. find
  24. Vasileios Hatzivassiloglou and Janyce Wiebe. 2000a. Effects of adjective orientation and gradability on sentence subjectivity. In COLING, pages 299–305.
  25. Vasileios Hatzivassiloglou and Janyce M Wiebe. 2000b. Effects of adjective orientation and gradability on sentence subjectivity. In Proceedings of the 18th conference on Computational linguistics-Volume 1, pages 299–305. Association for Computational Linguistics.
  26. Hochbaum and Shmoys. 1985. A best possible heuristic for the k-center problem. Mathematics of Operations Research, 10(2):180–184. ↗
  27. Hiroshi Kanayama and Tetsuya Nasukawa. 2006. Fully automatic lexicon expansion for domain-oriented sentiment analysis. In EMNLP’06, pages 355–363. find
  28. Michael Laver, Kenneth Benoit, and John Garry. 2003. Extracting policy positions from political texts using words as data. American Political Science Review, 97(02):311–331. ↗
  29. Wei-Hao Lin, Theresa Wilson, Janyce Wiebe, and Alexander Hauptmann. 2006. Which side are you on?: identifying perspectives at the document and sentence levels. In Proceedings of the Tenth Conference on Computational Natural Language Learning, pages 109–116. Association for Computational Linguistics. find
  30. Bing Liu and Lei Zhang. 2012. A survey of opinion mining and sentiment analysis. In Charu C. Aggarwal and ChengXiang Zhai, editors, Mining Text Data, pages 415–463. Springer US. ↗
  31. J. B. MacQueen. 1967. Some methods for classification and analysis of multivariate observations. In L. M. Le Cam and J. Neyman, editors, Proc. of the fifth Berkeley Symposium on Mathematical Statistics and Probability, volume 1, pages 281–297. University of California Press. find
  32. Christopher D. Manning, Prabhakar Raghavan, and Hinrich Schtze. 2008. Introduction to Information Retrieval. Cambridge University Press, New York, NY, USA. ↗
  33. Bo Pang and Lillian Lee. 2008. Opinion mining and sentiment analysis. Foundations and trends in information retrieval, 2(1-2):1–135. ↗
  34. Ana-Maria Popescu and Orena Etzioni. 2007. Extracting product features and opinions from reviews. In Natural language processing and text mining, pages 9–28. Springer. ↗
  35. Ellen Riloff and Janyce Wiebe. 2003. Learning extraction patterns for subjective expressions. In EMNLP’03, pages 105–112.
  36. Ellen Riloff, Siddharth Patwardhan, and Janyce Wiebe. 2006. Feature subsumption for opinion analysis. In Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, pages 440–448. Association for Computational Linguistics. find
  37. Swapna Somasundaran and Janyce Wiebe. 2009. Recognizing stances in online debates. In Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language Processing of the AFNLP, pages 226–234, Suntec, Singapore, August. Association for Computational Linguistics. find
  38. Hiroya Takamura, Takashi Inui, and Manabu Okumura. 2005. Extracting semantic orientations of words using spin model. In ACL’05, pages 133–140.
  39. Peter Turney and Michael Littman. 2003. Measuring praise and criticism: Inference of semantic orientation from association. ACM Transactions on Information Systems, 21:315–346. ↗
  40. Janyce Wiebe. 2000. Learning subjective adjectives from corpora. In Proceedings of the Seventeenth National Conference on Artificial Intelligence and Twelfth Conference on Innovative Applications of Artificial Intelligence, pages 735–740. find
  41. Theresa Wilson, Janyce Wiebe, and Paul Hoffmann. 2005. Recognizing contextual polarity in phrase-level sentiment analysis. In HLT/EMNLP’05, Vancouver, Canada.
  42. Zhongwu Zhai, Bing Liu, Hua Xu, and Peifa Jia. 2010. Grouping product features using semi-supervised learning with soft-constraints. In Proceedings of the 23rd International Conference on Computational Linguistics, pages 1272–1280. Association for Computational Linguistics. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/jbara-2013-identifying-opinion-subgroups-arabic-online/