---
title: "SAMAR — paper by Abdul-Mageed et al. (2012)"
description: "The paper focuses on Arabic social media text, classifying individual sentences by subjectivity and… 27 references, cited by 20 papers here."
authors: ["Muhammad Abdul-Mageed", "Sandra Kuebler", "Mona Diab"]
year: 2012
venue: "Proceedings of the 3rd Workshop in Computational Approaches to Subjectivity and Sentiment Analysis"
source: "https://aclanthology.org/W12-3705/"
licence: "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)"
licence_url: "https://creativecommons.org/licenses/by-nc-sa/3.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/mageed-2012-samar-system-subjectivity-sentiment-analysis/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/mageed-2012-samar-system-subjectivity-sentiment-analysis/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2012](https://md.research.elamur.ai/free/en/papers/2012/) / SAMAR: A System for Subjectivity and Sentiment Analysis of Arabic Social Media

# SAMAR: A System for Subjectivity and Sentiment Analysis of Arabic Social Media

**Authors:** [Muhammad Abdul-Mageed](https://md.research.elamur.ai/free/en/authors/muhammad-abdul-mageed/), [Sandra Kuebler](https://md.research.elamur.ai/free/en/authors/sandra-kuebler/) and [Mona Diab](https://md.research.elamur.ai/free/en/authors/mona-diab/) · **Year:** 2012 · **Venue:** Proceedings of the 3rd Workshop in Computational Approaches to Subjectivity and Sentiment Analysis · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY-NC-SA 3.0 (ACL Anthology, before 2016)](https://creativecommons.org/licenses/by-nc-sa/3.0/)

**Links:** [Publisher's page](https://aclanthology.org/W12-3705/) · [PDF](https://aclanthology.org/W12-3705.pdf) · [OpenAlex](https://openalex.org/W192118262)

**Citations:** 95 citations (OpenAlex, 2026-10-14) · [20](https://md.research.elamur.ai/free/en/papers/mageed-2012-samar-system-subjectivity-sentiment-analysis/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The paper focuses on Arabic social media text, classifying individual sentences by subjectivity and sentiment (SSA). Arabic is morphologically rich, and these genres mix Modern Standard Arabic (MSA) with dialectal Arabic (DA). Earlier SSA work mostly covered English and edited genres such as news and reviews.

The authors built four annotated corpora: DARDASHA (chat), TAGREED (Twitter), TAHRIR (Wikipedia Talk Pages) and MONTADA (Web forums). They labeled them for subjectivity and polarity and added metadata such as gender, user ID, document ID and language variety. SAMAR runs two cascaded binary SVM light classifiers with a linear kernel over binary presence features. The first separates objective from subjective sentences, and the second separates positive from negative ones.

The experiments vary four things:
- lexical representation (tokens or lemmas from AMIRA)
- POS tagsets (RTS, ERTS)
- standard features (a UNIQUE replacement for rare words and an MSA adjective polarity lexicon)
- dialect and genre features

No single configuration wins everywhere, but lemmatization appears in all the best setups. This shows that each domain and task needs its own tuning.

## Key points
- Corpus sizes: DAR has 2,798 chat turns, TGRD 3,015 tweets (1,466 MSA and 1,549 DA), THR 3,008 sentences and MONT 3,097 sentences. Two native-speaker annotators reached Kappa between 0.82 and 0.89.
- The polarity lexicon has 3,982 manually labeled newswire adjectives (positive, negative or neutral) taken from the Penn Arabic Treebank. AMIRA, the preprocessing tool, is trained only on MSA newswire.
- Best subjectivity accuracy was 95.83 on DAR, 72.52 on TGRD, 83.33 on THR and 84.19 on MONT. Best sentiment accuracy was 71.28, 65.87, 67.44 and 81.36 on the same corpora. All best subjectivity settings except MONT include ERTS tags.
- Adding an explicit MSA/DA variety feature to the full TGRD corpus did not beat the best setting without it (72.52 vs 69.54 for subjectivity).
- Genre metadata never helped subjectivity, but gender and user ID improved sentiment. Accuracy rose by 1.73 points on DAR and 4.07 on THR. The document ID feature was useless.

## Interesting ideas
- The paper directly compares two ways of reducing word forms, keeping clitic-split tokens versus collapsing to citation lemmas, and lemmas consistently win against sparsity in short texts.
- Subjectivity is decided first, and polarity is judged only for sentences already found subjective, so each binary decision stays simple and inspectable.
- POS tagsets with controlled amounts of morphology (RTS vs ERTS) let the authors measure how much inflectional detail actually helps a task.
- The best-setting tables show very high accuracy alongside an F-score of 0.00 for the objective class, so a skewed classifier can look strong if only accuracy is reported.
- Speaker attributes such as gender and author identity helped polarity but not subjectivity, which suggests that who is speaking shapes how they evaluate.

## For Mova

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

- **I1** (tokens): We could test our lemmatizer against downstream tasks rather than in isolation, measuring how often lemma-level features beat surface forms in frame and role matching.
- **I2** (frames): A speech-act layer could first mark a clause as reporting versus evaluating, then assign stance or irony only to evaluative clauses, which keeps each decision traceable.
- **I3** (tokens): We could ablate morphological features in our tagger output one at a time to see which ones the parser and frame builder really depend on.
- **I4** (learn): Evaluation reports should always include per-class F and a majority-class baseline as a negative control, so degenerate models are caught automatically.
- **I5** (world): We could store the speaker entity and its known traits in the text's world so stance interpretation can condition on the opinion holder rather than on words alone.

*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

- [Columbia University](https://md.research.elamur.ai/free/en/universities/columbia-university/) — University, New York, United States · 241 works · 54,053 citations
- [Indiana University Bloomington](https://md.research.elamur.ai/free/en/universities/indiana-university-bloomington/) — University, Bloomington, United States · 37 works · 3,685 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [💭 Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/) · 1,552 works · [💭 Experimental and Cognitive Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/experimental-and-cognitive-psychology/) · 870 works

🧭 Topics (2): [😊 Sentiment & emotion](https://md.research.elamur.ai/free/en/topics/sentiment-emotion/) · 972 works · [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 2,908 works

## Cite

Cite as: Muhammad Abdul-Mageed, Sandra Kuebler and Mona Diab. 2012. SAMAR: A System for Subjectivity and Sentiment Analysis of Arabic Social Media. Proceedings of the 3rd Workshop in Computational Approaches to Subjectivity and Sentiment Analysis.

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

```bibtex
@inproceedings{abdulmageed2012samar,
  title = {{SAMAR: A System for Subjectivity and Sentiment Analysis of Arabic Social Media}},
  author = {Muhammad Abdul-Mageed, Sandra Kuebler and Mona Diab},
  booktitle = {Proceedings of the 3rd Workshop in Computational Approaches to Subjectivity and Sentiment Analysis},
  year = {2012},
  url = {https://aclanthology.org/W12-3705/}
}
```

- [← previous Embedding epistemic modals in English: A corpus-based study](https://md.research.elamur.ai/free/en/papers/hacquard-2012-embedding-epistemic-modals-english-corpus/)
- [next → Cross-Genre and Cross-Domain Detection of Semantic Uncertainty](https://md.research.elamur.ai/free/en/papers/szarvas-2012-cross-genre-cross-domain-detection/)
- [↑ back to papers of 2012](https://md.research.elamur.ai/free/en/papers/2012/)

## References

27 references, 9 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 Transactions on Information Systems, 26:1–34. [find](https://scholar.google.com/scholar?q=Ahmed+Abbasi%2C+Hsinchun+Chen%2C+and+Arab+Salem%2E+2008%2E+Sentiment+analysis+in+multiple+languages%3A+Feature+selection+for+opinion+classification+in+Web+forums%2E+ACM+Transactions+on+Information+Systems%2C+26%3A1%E2%80%933)
2. Muhammad Abdul-Mageed and Mona Diab. 2012a. [AWATIF: A multi-genre corpus for Modern Standard Arabic subjectivity and sentiment analysis](https://md.research.elamur.ai/free/en/papers/mageed-2012-awatif-multi-genre-corpus-modern/). In Proceedings of LREC, Istanbul, Turkey.
3. 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](https://scholar.google.com/scholar?q=Muhammad+Abdul%2DMageed+and+Mona+Diab%2E+2012b%2E+Toward+building+a+large%2Dscale+Arabic+sentiment+lexicon%2E+In+Proceedings+of+the+6th+International+Global+Word%2DNet+Conference%2C+Matsue%2C+Japan%2E)
4. Muhammad Abdul-Mageed, Hamdan Albogmi, Abdulrahman Gerrio, Emhamed Hamed, and Omar Aldibasi. 2011a. Tweeting in Arabic: What, how and whither. Presented at the 12th Annual Conference of the Association of Internet Researchers (Internet Research 12.0, Performance and Participation), Seattle, WA. [find](https://scholar.google.com/scholar?q=Muhammad+Abdul%2DMageed%2C+Hamdan+Albogmi%2C+Abdulrahman+Gerrio%2C+Emhamed+Hamed%2C+and+Omar+Aldibasi%2E+2011a%2E+Tweeting+in+Arabic%3A+What%2C+how+and+whither%2E+Presented+at+the+12th+Annual+Conference+of+the+Associatio)
5. Muhammad Abdul-Mageed, Mona Diab, and Mohamed Korayem. 2011b. [Subjectivity and sentiment analysis of Modern Standard Arabic](https://md.research.elamur.ai/free/en/papers/mageed-2011-subjectivity-sentiment-analysis-modern-standard/). In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, pages 587–591, Portland, OR.
6. Muhammad Abdul-Mageed, Mohamed Korayem, and Ahmed YoussefAgha. 2011c. ”Yes we can?”: Subjectivity annotation and tagging for the health domain. In Proceedings of RANLP2011, Hissar, Bulgaria. [find](https://scholar.google.com/scholar?q=Muhammad+Abdul%2DMageed%2C+Mohamed+Korayem%2C+and+Ahmed+YoussefAgha%2E+2011c%2E+%E2%80%9DYes+we+can%3F%E2%80%9D%3A+Subjectivity+annotation+and+tagging+for+the+health+domain%2E+In+Proceedings+of+RANLP2011%2C+Hissar%2C+Bulgaria%2E)
7. Ann Banfield. 1982. Unspeakable Sentences: Narration and Representation in the Language of Fiction. Routledge, Boston. [find](https://scholar.google.com/scholar?q=Ann+Banfield%2E+1982%2E+Unspeakable+Sentences%3A+Narration+and+Representation+in+the+Language+of+Fiction%2E+Routledge%2C+Boston%2E)
8. Penelope Brown and Stephen Levinson. 1987. [Politeness](https://md.research.elamur.ai/free/en/papers/brown-1987-politeness-some-universals-language-usage/): Some Universals in Language Usage. Cambridge University Press.
9. Rebecca Bruce and Janyce Wiebe. 1999. Recognizing subjectivity. A case study of manual tagging. Natural Language Engineering, 5(2):187–205. [find](https://scholar.google.com/scholar?q=Rebecca+Bruce+and+Janyce+Wiebe%2E+1999%2E+Recognizing+subjectivity%2E+A+case+study+of+manual+tagging%2E+Natural+Language+Engineering%2C+5%282%29%3A187%E2%80%93205%2E)
10. Kushal Dave, Steve Lawrence, and David Pennock. 2003. Mining the peanut gallery: Opinion extraction and semantic classification of product reviews. In Proceedings of the 12th International Conference on World Wide Web, pages 519–528, Budapest, Hungary. ACM. [find](https://scholar.google.com/scholar?q=Kushal+Dave%2C+Steve+Lawrence%2C+and+David+Pennock%2E+2003%2E+Mining+the+peanut+gallery%3A+Opinion+extraction+and+semantic+classification+of+product+reviews%2E+In+Proceedings+of+the+12th+International+Conference+)
11. Mona Diab, Dan Jurafsky, and Kadri Hacioglu. 2007. Automatic processing of Modern Standard Arabic text. In Abdelhadi Soudi, Antal van den Bosch, and Günter Neumann, editors, Arabic Computational Morphology. Springer. [↗](https://doi.org/10.1007/978-1-4020-6046-5_9)
12. Mona Diab, Nizar Habash, Owen Rambow, Mohamed Altantawy, and Yassin Benajiba. 2010. COLABA: Arabic dialect annotation and processing. In LREC Workshop on Semitic Language Processing, pages 66– 74, Valetta, Malta. [find](https://scholar.google.com/scholar?q=Mona+Diab%2C+Nizar+Habash%2C+Owen+Rambow%2C+Mohamed+Altantawy%2C+and+Yassin+Benajiba%2E+2010%2E+COLABA%3A+Arabic+dialect+annotation+and+processing%2E+In+LREC+Workshop+on+Semitic+Language+Processing%2C+pages+66%E2%80%93+74%2C+Val)
13. Mona Diab. 2007a. [Improved Arabic base phrase chunking with a new enriched POS tag set](https://md.research.elamur.ai/free/en/papers/diab-2007-improved-arabic-base-phrase-chunking/). In Proceedings of the 2007 Workshop on Computational Approaches to Semitic Languages: Common Issues and Resources, pages 89–96, Prague, Czech Republic.
14. Mona Diab. 2007b. Towards an optimal POS tag set for Modern Standard Arabic processing. In Proceedings of Recent Advances in Natural Language Processing (RANLP), Borovets, Bulgaria. [find](https://scholar.google.com/scholar?q=Mona+Diab%2E+2007b%2E+Towards+an+optimal+POS+tag+set+for+Modern+Standard+Arabic+processing%2E+In+Proceedings+of+Recent+Advances+in+Natural+Language+Processing+%28RANLP%29%2C+Borovets%2C+Bulgaria%2E)
15. Mona Diab. 2009. Second generation AMIRA tools for Arabic processing: Fast and robust tokenization, POS tagging, and base phrase chunking. In Proceedings of the Second International Conference on Arabic Language Resources and Tools, pages 285–288, Cairo, Egypt. [find](https://scholar.google.com/scholar?q=Mona+Diab%2E+2009%2E+Second+generation+AMIRA+tools+for+Arabic+processing%3A+Fast+and+robust+tokenization%2C+POS+tagging%2C+and+base+phrase+chunking%2E+In+Proceedings+of+the+Second+International+Conference+on+Arab)
16. Nizar Habash, Owen Rambow, and Ryan Roth. 2009. MADA+TOKAN: A toolkit for Arabic tokenization, diacritization, morphological disambiguation, POS tagging, stemming and lemmatization. In Proceedings of the Second International Conference on Arabic Language Resources and Tools, pages 102–109, Cairo, Egypt. [find](https://scholar.google.com/scholar?q=Nizar+Habash%2C+Owen+Rambow%2C+and+Ryan+Roth%2E+2009%2E+MADA%2BTOKAN%3A+A+toolkit+for+Arabic+tokenization%2C+diacritization%2C+morphological+disambiguation%2C+POS+tagging%2C+stemming+and+lemmatization%2E+In+Proceedings+of+)
17. Susan Herring. 1996. Bringing familiar baggage to the new frontier: Gender differences in computer-mediated communication. In J. Selzer, editor, Conversations. Allyn & Bacon. [find](https://scholar.google.com/scholar?q=Susan+Herring%2E+1996%2E+Bringing+familiar+baggage+to+the+new+frontier%3A+Gender+differences+in+computer%2Dmediated+communication%2E+In+J%2E+Selzer%2C+editor%2C+Conversations%2E+Allyn+%26+Bacon%2E)
18. Minqing Hu and Bing Liu. 2004. Mining and summarizing customer reviews. In Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 168–177, Seattle, WA. [find](https://scholar.google.com/scholar?q=Minqing+Hu+and+Bing+Liu%2E+2004%2E+Mining+and+summarizing+customer+reviews%2E+In+Proceedings+of+the+Tenth+ACM+SIGKDD+International+Conference+on+Knowledge+Discovery+and+Data+Mining%2C+pages+168%E2%80%93177%2C+Seattle%2C+)
19. Thorsten Joachims. 2008. Svmlight: Support vector machine. [http://svmlight.joachims.org/](http://svmlight.joachims.org/), [↗](http://svmlight.joachims.org/)
20. Cornell University, 2008. [find](https://scholar.google.com/scholar?q=Cornell+University%2C+2008%2E)
21. Soo-Min Kim and Eduard Hovy. 2004. [Determining the sentiment of opinions](https://md.research.elamur.ai/free/en/papers/kim-2004-determining-sentiment-opinions/). In Proceedings of the 20th International Conference on Computational Linguistics, pages 1367–1373, Geneva, Switzerland.
22. Mohamed Maamouri, Anne Bies, Tim Buckwalter, and W. Mekki. 2004. The Penn Arabic Treebank: Building a large-scale annotated Arabic corpus. In NEM-LAR Conference on Arabic Language Resources and Tools, pages 102–109, Cairo, Egypt. [find](https://scholar.google.com/scholar?q=Mohamed+Maamouri%2C+Anne+Bies%2C+Tim+Buckwalter%2C+and+W%2E+Mekki%2E+2004%2E+The+Penn+Arabic+Treebank%3A+Building+a+large%2Dscale+annotated+Arabic+corpus%2E+In+NEM%2DLAR+Conference+on+Arabic+Language+Resources+and+Tools%2C)
23. Peter Turney. 2002. [Thumbs up or thumbs down? Semantic orientation applied to unsupervised classification of reviews](https://md.research.elamur.ai/free/en/papers/turney-2002-thumbs-up-or-thumbs-down/). In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics (ACL’02), Philadelphia, PA.
24. Janyce Wiebe, Rebecca Bruce, and Tim O’Hara. 1999. [Development and use of a gold standard data set for subjectivity classifications](https://md.research.elamur.ai/free/en/papers/wiebe-1999-development-use-gold-standard-data/). In Proceedings of the 37th Annual Meeting of the Association for Computational Linguistics (ACL-99), pages 246–253, University of Maryland.
25. Janyce Wiebe, Theresa Wilson, Rebecca Bruce, Matthew Bell, and Melanie Martin. 2004. [Learning subjective language](https://md.research.elamur.ai/free/en/papers/wiebe-2004-learning-subjective-language/). Computational Linguistics, 30:227–308. [↗](https://doi.org/10.1162/0891201041850885)
26. Jeonghee Yi, Tetsuya Nasukawa, Razvan Bunescu, and Wayne Niblack. 2003. Sentiment analyzer: Extracting sentiments about a given topic using natural language processing techniques. In Proceedings of the 3rd IEEE International Conference on Data Mining, pages 427–434, Melbourne, FL. [find](https://scholar.google.com/scholar?q=Jeonghee+Yi%2C+Tetsuya+Nasukawa%2C+Razvan+Bunescu%2C+and+Wayne+Niblack%2E+2003%2E+Sentiment+analyzer%3A+Extracting+sentiments+about+a+given+topic+using+natural+language+processing+techniques%2E+In+Proceedings+of+th)
27. Hong Yu and Vasileios Hatzivassiloglou. 2003. [Towards answering opinion questions: Separating facts from opinions and identifying the polarity of opinion sentences](https://md.research.elamur.ai/free/en/papers/yu-2003-answering-opinion-questions-separating-facts/). In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), Sapporo, Japan. [↗](https://doi.org/10.7916/d88w3nn0)

*Updated 1 November 2026*