Elamur Research⌕
☰
home / papers / 2015 / Sentiment Analysis For Modern Standard Arabic And Colloquial

Sentiment Analysis For Modern Standard Arabic And Colloquial

Paper · 2015 · Hossam S. Ibrahim, Sherif M. Abdou and Mervat Gheith

Summary

The paper tackles sentence-level sentiment polarity classification for Arabic. This is hard because online text mixes Modern Standard Arabic (MSA) with Egyptian colloquial dialect, and earlier Arabic systems used few features, often ignored negation, and lacked dialect coverage. The authors take a semi-supervised, feature-based route built on two resources:

  • Polarity lexicon: an Arabic sentiment-word lexicon that grows automatically. It starts from a hand-collected, hand-annotated gold seed. New words are added and their orientation inferred by aggregating synsets from free online Arabic dictionaries and thesauri, guided by POS tags and synonym/antonym relations.
  • Idioms lexicon: a dedicated lexicon of Arabic idioms and sayings that carry sentiment.

On top of these, an SVM classifier uses linguistically motivated features: context-sensitive intensifiers, context-sensitive shifters, negation handling, and syntactic cues for sentences containing conflicting phrases. The data are MSA and Egyptian dialect tweets and microblogs from domains such as hotel reservations and product reviews. The reported accuracy exceeds 95%, and the authors separately measure how much lexicon expansion and the idioms lexicon contribute. The work matters because it gives dialect-aware resources and shows that fixed multiword expressions are a major sentiment signal.

Key points

  • The task is sentence-level polarity on MSA plus Egyptian dialect tweets and microblogs (hotel reservations, product reviews, etc.).
  • The classifier is an SVM (the paper cites SVM-light), and the reported accuracy is above 95%.
  • The polarity lexicon grows from a manually annotated gold seed. It expands via synset aggregation over free online Arabic lexicons and thesauri, using POS tags and synonym/antonym links. A companion CICLing 2015 paper describes this lexicon.
  • The idioms/sayings lexicon draws on printed collections of Egyptian popular sayings and the PROz online glossary of Arabic idioms.
  • The feature set includes contextual intensifiers, contextual shifters, negation handling, and syntactic features for conflicting phrases. Ablations show the gains from lexicon expansion and from the idioms lexicon.

Interesting ideas

  • I1. Treat whole idioms and sayings as single sentiment units with their own lexicon, because their polarity cannot be composed from the individual words.
  • I2. Grow a polarity lexicon from a small trusted seed by voting across the synsets of several free dictionaries, with antonym links flipping the sign, so each new entry has a traceable derivation.
  • I3. Model intensifiers and shifters as context-dependent operators that change the strength or direction of a nearby sentiment word, rather than as standalone polarity carriers.
  • I4. Give sentences with conflicting positive and negative phrases explicit syntactic features, so the classifier can tell which clause dominates.
  • I5. Run ablations that isolate each resource (lexicon expansion, idioms lexicon) to show its separate contribution to accuracy.

For Mova

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

  • I1 (frames): Add a lexicon of evaluative English idioms matched over UD subtrees, so prag assigns stance from the fixed expression rather than from misleading literal words.
  • I2 (world): Extend compiled value and evaluation knowledge from WordNet synonym/antonym links, storing the derivation path so every inherited polarity stays explainable.
  • I3 (frames): Implement intensity and negation scope as operators attached along dependency arcs, so frames record which modifier changed an evaluation and why.
  • I4 (tree): Use UD clause relations such as advcl, conj, and contrastive markers to decide which clause's evaluation governs the sentence, keeping the decision rule-visible.
  • I5 (learn): Evaluate each compiled resource with on/off ablations as a negative control, so additions are kept only when they measurably help.

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

🎓Cairo UniversityGiza · Egypt43 works · 1,352 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 2Research subjects and languages the work is about.Sentiment & emotion · 1,035 worksSemitic languages · 3,066 works

Cite

Cite as Hossam S. Ibrahim, Sherif M. Abdou and Mervat Gheith. 2015. Sentiment Analysis For Modern Standard Arabic And Colloquial. arXiv:1505.03105.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.5121/ijnlc.2015.4207 · arXiv 1505.03105

@misc{ibrahim2015sentiment,
  title = {{Sentiment Analysis For Modern Standard Arabic And Colloquial}},
  author = {Hossam S. Ibrahim, Sherif M. Abdou and Mervat Gheith},
  year = {2015},
  doi = {10.5121/ijnlc.2015.4207},
  eprint = {1505.03105},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.5121/ijnlc.2015.4207}
}

References

40 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. A. Shoukry and A. Rafea, "Sentence-level Arabic sentiment analysis," in Collaboration Technologies and Systems (CTS) International Conference, Denver, CO, USA, 2012, pp. 546-550. find
  2. B. Pang, L. Lee, and S. Vaithyanathan, "Thumbs up? Sentiment classification using machine learning techniques," in Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), 2002, pp. 79–86. ↗
  3. D. Davidiv, O. Tsur, and A. Rappoport, "Enhanced Sentiment Learning Using Twitter Hash-tags and Smileys," in Proceedings of the 23rd International Conference on Computational Linguistics (Coling2010), Beijing, China, 2010, pp. 241–249. find
  4. L. Barbosa and J. Feng, "Robust Sentiment Detection on Twitter from Biased and Noisy Data " in Proceedings of the 23rd International Conference on Computational Linguistics (Coling), 2010.
  5. P. Turney, "Thumbs Up or Thumbs Down? Semantic Orientation Applied to Unsupervised Classification of Reviews," in Proceedings of the 40th Annual Meeting on Association for Computational Linguistics ACL '02, Stroudsburg, PA, USA, 2002, pp. 417-424. ↗
  6. V. Hatzivassiloglou and K. R. McKeown, "Predicting the semantic orientation of adjectives," in Proceedings of the Joint ACL / EACL Conference, 1997, pp. 174–181. find
  7. B. Pang and L. Lee, "Opinion mining and sentiment analysis," Foundations and Trends in Information Retrieval vol. 2, pp. 1–135, 2008. ↗
  8. M. Hu and B. Liu, "Mining and summarizing customer reviews " in Proceedings of the ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2004, pp. 168–177. find
  9. B. Liu, "Sentiment Analysis and Subjectivity," in Handbook of Natural Language Processing, Second ed: CRC Press, Taylor and Francis Group, 2010. find
  10. P. Alexander and P. Patrick, "Twitter as a Corpus for Sentiment Analysis and Opinion Mining " in Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10), European Language Resources Association ELRA, Valletta, Malta, 2010.
  11. C. Scheible and H. Schütze, "Bootstrapping Sentiment Labels For Unannotated Documents With Polarity PageRank," in Proceedings of the Eight International Conference on Language Resources and Evaluation (LREC 2012), Istambol-Turki, 2012. find
  12. C. Manning and D. Klein, "Optimization, maxent models, and conditional estimation without magic," in Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology, 2003, p. 8. find
  13. A. Abbasi, H. Chen, and A. Salem, "Sentiment Analysis in Multiple Languages: Feature Selection for Opinion Classification in Web Forums," ACM Transactions on Information Systems, vol. 26, 2008. find
  14. E. Riloff and J. Wiebe, "Learning extraction patterns for subjective expressions," in Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), 2003.
  15. E. Riloff, J. Wiebe, and T. Wilson, "Learning subjective nouns using extraction pattern bootstrapping," in Proceedings of the Conference on Natural Language Learning (CoNLL), 2003, pp. 25–32.
  16. M. Abdul-Mageed and M. Diab, "Subjectivity and Sentiment Annotation of Modern Standard Arabic Newswire," in Proceedings of the Fifth Law Workshop (LAW V), Association for Computational Linguistics, Portland, Oregon, 2011, pp. 110–118.
  17. M. Abdul-Mageed, M. Diab, and M. Korayem, "Subjectivity and sentiment analysis of modern standard Arabic," in Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics, 2011.
  18. M. Abdul-Mageed, K. Sandra, and M. Diab, "SAMAR: A System for Subjectivity and Sentiment Analysis of Arabic Social Media," in Proceedings of the 3rd Workshop on Computational Approaches to Subjectivity and Sentiment Analysis, Jeju,Republic of Korea, 2012, pp. 19–28.
  19. A. Mourad and K. Darwish, "Subjectivity and Sentiment Analysis of Modern Standard Arabic and Arabic Microblogs," in Proceedings of the 4th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis (WASSA), Atlanta, Georgia, 2013, pp. 55–64.
  20. M. Korayem, D. Crandall, and M. Abdul-Mageed, "Subjectivity and Sentiment Analysis of Arabic: A Survey," in Advanced Machine Learning Technologies and Applications, Communications in Computer and Information Science series 322, (Springer), AMLTA, 2012, pp. 128-139. find
  21. M. Abdul-Mageed and M. Diab, "AWATIF: A multi-genre corpus for Arabic subjectivity and sentiment analysis," in Proceedings of the 8th International Conference on Language Resources and Evaluation (LREC), Istanbul, Turkey, 2012a. find
  22. M. Rushdi-Saleh, M. Mart´ın-Valdivia, L. Ure˜na-L´opez, and J. Perea-Ortega, "Oca: Opinion corpus for Arabic," Journal of the American Society for Information Science and Technology, vol. 62, pp. 2045–2054, 2011. ↗
  23. M. Elarnaoty, S. AbdelRahman, and A. Fahmy, "A Machine Learning Approach for Opinion Holder Extraction Arabic Language," CoRR, abs/1206.1011, vol. 3, 2012. ↗
  24. M. Abdul-Mageed and M. Diab, "SANA: A Large Scale Multi-Genre, Multi-Dialect Lexicon for Arabic Subjectivity and Sentiment Analysis," in Proceedings of The 9th edition of the Language Resources and Evaluation Conference (LREC ), Reykjavik, Iceland, 2014.
  25. E. Refaee and V. Rieser, "An Arabic Twitter Corpus for Subjectivity and Sentiment Analysis," in Proceedings of The 9th edition of the Language Resources and Evaluation Conference (LREC 2014), Reykjavik, Iceland, 2014.
  26. M. Elmahdy, G. Rainer, M. Wolfgang, and A. Slim, "Survey on common Arabic language forms from a speech recognition point of view," in proceeding of International conference on Acoustics (NAG-DAGA), Rotterdam, Netherlands, 2009, pp. 63-66. find
  27. J. C. Carletta, "Assessing agreement on classification tasks: the KAPPA statistic " Computational Linguistics, vol. 22, pp. 249- 254, 1996. ↗
  28. B. Liu, Sentiment Analysis and Opinion Mining Morgan &Claypool Publishers, 2012. find
  29. A. Basha, ىعوضوم فاشك عم لثملا نم لولاا فرحلا بسح ةبترمو ةحورشم : ةيملاعلا لاثملاا [Colloquial sayings: an annotated and arranged by the first letter of ideals with the Scout TOPICAL]. Egypt: Al-Ahram Foundation - Al-Ahram Center for Translation and Publishing, 1986. find
  30. A. Saalan, ةيرصملا ةيبعشلا لاثملاا ةعوسوم [Encyclopedia of Egyptian popular sayings], First ed. Egypt: Dar-alafkalarabia press, 2003. find
  31. F. Husain, ىرصملا رولكلوفلا , ةيماعلا لاثملاا , ةيبعشلا صصقلا , ةيبرعلا رداونلا [Colloquial sayings, Egyptian folklore]. Egypt: General Egyptian Book Organization GEBO, 1984. find
  32. G. Taher. (2006). ةيملع ةسارد - ةيبعشلا لاثملأا ةعوسوم [Encyclopedia of public sayings - a scientific study]. Available: http://books.google.com.eg/books?id=2CR\_EKTjxRgC ↗
  33. PROz. (2014). PROz website for Arabic Idioms/Maxims/Sayings (Jan 2014). Available: http://www.proz.com/glossary-translations/ ↗
  34. M. Diab, "Towards an optimal POS tag set for Modern Standard Arabic processing," in Proceedings of Recent Advances in Natural Language Processing (RANLP), Borovets, Bulgaria, 2007. find
  35. O. F. Zaidan and C. Callison-Burch, "Arabic dialect identification," Computational Linguistics, vol. 40, pp. 171-202, March 2014 2012. ↗
  36. H. S. Ibrahim, S. M. Abdou, and M. Gheith, "Automatic expandable large-scale sentiment lexicon of Modern Standard Arabic and Colloquial," in 16th International Conference on Intelligent Text Processing and Computational Linguistics (CICLING), Cairo - Egypt, 2015. find
  37. M. Sharifi and W. Cohen. (2008, May, 2014). “Finding domain specifc polar words for sentiment classification. Available: http://www.cs.cmu.edu/~mehrbod/polarity_08.pdf ↗
  38. J. YI, T. NASUKAWA, R. BUNESCU, and W. NIBLACK, "Sentiment analyzer: Extracting sentiments about a given topic using natural language processing techniques " in Proceedings of the 3rd IEEE International Conference on Data Mining (ICDM), 2003, pp. 427–434. find
  39. Z. Fei, J. LIU, and G. WU, "Sentiment classification using phrase patterns," in Proceedings of the 4th IEEE International Conference on Computer Information Technology, 2004, pp. 1147–1152. find
  40. T. Joachims. (2008, Jan-2013). SVM-light: Support vector machine. Available: http://svmlight.joachims.org/ ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/ibrahim-2015-sentiment-analysis-modern-standard-arabic/