---
title: "Ambiguous Arabic Words Disambiguation: The Results (2009)"
description: "The paper addresses word sense disambiguation (WSD) for Arabic, where vowel marks are usually absent, words carry… 15 references."
authors: ["Laroussi Merhbene", "Anis Zouaghi", "Mounir Zrigui"]
year: 2009
venue: "Proceedings of the Student Research Workshop"
source: "https://aclanthology.org/R09-2009/"
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/merhbene-2009-ambiguous-arabic-words-disambiguation-results/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/merhbene-2009-ambiguous-arabic-words-disambiguation-results/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2009](https://md.research.elamur.ai/free/en/papers/2009/page/2/) / Ambiguous Arabic Words Disambiguation: The Results

# Ambiguous Arabic Words Disambiguation: The Results

**Authors:** [Laroussi Merhbene](https://md.research.elamur.ai/free/en/authors/laroussi-merhbene/), [Anis Zouaghi](https://md.research.elamur.ai/free/en/authors/anis-zouaghi/) and [Mounir Zrigui](https://md.research.elamur.ai/free/en/authors/mounir-zrigui/) · **Year:** 2009 · **Venue:** [Proceedings of the Student Research Workshop](https://md.research.elamur.ai/free/en/venues/ranlp/) ([RANLP 2009](https://md.research.elamur.ai/free/en/venues/ranlp/2009/)) · **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/R09-2009/) · [PDF](https://aclanthology.org/R09-2009.pdf) · [OpenAlex](https://openalex.org/W2250221192)

**Citations:** 4 citations (OpenAlex, 2026-10-14)

## Summary
The paper addresses word sense disambiguation (WSD) for Arabic, where vowel marks are usually absent, words carry attached clitics, and labelled resources are scarce. The work is part of a spoken-Arabic understanding pipeline. Because annotated data is lacking, the authors use an unsupervised, retrieval-style approach.

They gathered newspaper texts from the web across several domains. A linguist identified ambiguous words and assigned senses to their contexts. The pool of contexts for each sense was enlarged with contexts of synonyms taken from Arabic WordNet.

Each sense is represented by a compact "signature" of high tf × idf words. Signatures are optionally reduced to roots and tagged with the Brill tagger. Four IR similarity measures (LSA, Harman, Croft, Okapi) compare the current context with each sense's contexts, and the Lesk algorithm arbitrates when the measures disagree.

On 10 ambiguous words the full system reaches 73% correct senses, compared with 52% on raw contexts. The result shows that classic IR machinery can give usable WSD for a low-resource language.

## Key points
- The test set has 10 ambiguous Arabic words, such as عين (ayn), whose senses include eye, water spring, spy, "same thing" and the letter name. Sense definitions come from the Al Wasit dictionary.
- Each sense context is reduced to its top 20 tf × idf words, which also filters out stop words. Synonyms from Arabic WordNet supply extra contexts for each sense.
- The ablations show these accuracies:
  - Raw contexts: 52%.
  - Root normalisation: 56%.
  - Syntactic tags (noun, verb or particle from the Brill tagger, which itself is 78% accurate): 64.3%.
  - Signatures plus tags, combined via Lesk: 73%.
- Context similarity is computed from the weights of words shared by the two contexts, normalised. The weights come from the Harman, Croft or Okapi formulas.
- The authors note that comparison with other work is hard. Original Lesk is cited at 50–70%, and Karov & Edelman at 92% on four words. Future work proposes a multi-agent selector over the methods.

## Interesting ideas
- Run several different similarity measures side by side and call an independent arbitrator (here Lesk) only when they disagree, instead of trusting any single scorer.
- Grow the evidence for a rare sense by pooling contexts of its synonyms, so the sense profile does not depend on a handful of examples.
- Represent each sense by a short list of its most distinctive context words, so a decision can be explained by showing which signature words matched.
- Key context words by their category as well as their surface form; adding even a coarse three-way POS distinction clearly raised accuracy.
- Measure every preprocessing step separately against a plain baseline (raw, normalised, tagged, full) so the contribution of each step is visible.

## For Mova

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

- **I1** (learn): The same pattern could resolve disagreements among witness parsers when building silver data, with a separate, interpretable judge deciding the contested arcs.
- **I2** (world): Contexts of all WordNet synset members could be pooled into one profile per synset, giving sparse senses in the ontology usable distributional evidence.
- **I3** (frames): VerbNet class or frame choice for a polysemous verb could rest on small signature lists over UD dependents, keeping each role assignment traceable.
- **I4** (tokens): Signature and lexicon entries could use lemma+UPOS keys, which the tagger already produces, to reduce false matches between homographs.
- **I5** (learn): Mova's evaluation could include such stepwise ablations alongside negative controls, to show which pipeline stage actually earns its gains.

*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 Monastir](https://md.research.elamur.ai/free/en/universities/university-of-monastir/) — University, Monastir, Tunisia · 8 works · 52 citations
- [University of Sousse](https://md.research.elamur.ai/free/en/universities/university-of-sousse/) — University, Sousse, Tunisia · 5 works · 23 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works

🧭 Topics (1): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 3,066 works

## Cite

Cite as: Laroussi Merhbene, Anis Zouaghi and Mounir Zrigui. 2009. Ambiguous Arabic Words Disambiguation: The Results. Proceedings of the Student Research Workshop.

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

```bibtex
@inproceedings{merhbene2009ambiguous,
  title = {{Ambiguous Arabic Words Disambiguation: The Results}},
  author = {Laroussi Merhbene, Anis Zouaghi and Mounir Zrigui},
  booktitle = {Proceedings of the Student Research Workshop},
  year = {2009},
  url = {https://aclanthology.org/R09-2009/}
}
```

- [← previous A Study of a Segmentation Technique for Dialogue Act Assignation (short paper)](https://md.research.elamur.ai/free/en/papers/hinarejos-2009-study-segmentation-technique-dialogue-act/)
- [next → Arabic Disambiguation Using Dependency Grammar](https://md.research.elamur.ai/free/en/papers/daoud-2009-arabic-disambiguation-dependency-grammar/)
- [↑ back to papers of 2009](https://md.research.elamur.ai/free/en/papers/2009/page/2/)

## References

15 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Black, W. J. & Elkateb, S. (2004) A Prototype English-Arabic Dictionary Based on WordNet, Proceedings of 2nd Global WordNet Conference, GWC2004, Czech Republic: 67-74. [find](https://scholar.google.com/scholar?q=Black%2C+W%2E+J%2E+%26+Elkateb%2C+S%2E+%282004%29+A+Prototype+English%2DArabic+Dictionary+Based+on+WordNet%2C+Proceedings+of+2nd+Global+WordNet+Conference%2C+GWC2004%2C+Czech+Republic%3A+67%2D74%2E)
2. BRILL E. (1993), A Corpus-Based Approach to Language Learning, Thesis non-published, University of Pennsylvania, Department of Computer and Information Science. [find](https://scholar.google.com/scholar?q=BRILL+E%2E+%281993%29%2C+A+Corpus%2DBased+Approach+to+Language+Learning%2C+Thesis+non%2Dpublished%2C+University+of+Pennsylvania%2C+Department+of+Computer+and+Information+Science%2E)
3. Croft.W, 1983. Experiments with representation in a document retrieval system; Research and development, 2(1) ; pp. 1-21 ; 1983. [find](https://scholar.google.com/scholar?q=Croft%2EW%2C+1983%2E+Experiments+with+representation+in+a+document+retrieval+system%3B+Research+and+development%2C+2%281%29+%3B+pp%2E+1%2D21+%3B+1983%2E)
4. De Loupy, 2000. Assessing the contribution of linguistic knowledge in semantic disambiguation and information retrieval. THESIS presented in the University of Avignon and the country of Vaucluse. [find](https://scholar.google.com/scholar?q=De+Loupy%2C+2000%2E+Assessing+the+contribution+of+linguistic+knowledge+in+semantic+disambiguation+and+information+retrieval%2E+THESIS+presented+in+the+University+of+Avignon+and+the+country+of+Vaucluse%2E)
5. Derwester, S., Dumais, S.T., Furnas, G.W., Landauer, T.K., Harshmann, R. 1990. Indexing by Latent Semantic Analysis. Journal of the American Society for Informartion Science, 41 : 391-407. [find](https://scholar.google.com/scholar?q=Derwester%2C+S%2E%2C+Dumais%2C+S%2ET%2E%2C+Furnas%2C+G%2EW%2E%2C+Landauer%2C+T%2EK%2E%2C+Harshmann%2C+R%2E+1990%2E+Indexing+by+Latent+Semantic+Analysis%2E+Journal+of+the+American+Society+for+Informartion+Science%2C+41+%3A+391%2D407%2E)
6. Dictionary Al Wasit, 4th Edition, 2003, Academy of the Arabic language, Sunrise International Library. [find](https://scholar.google.com/scholar?q=Dictionary+Al+Wasit%2C+4th+Edition%2C+2003%2C+Academy+of+the+Arabic+language%2C+Sunrise+International+Library%2E)
7. Florentina Vasilescu , 2003. Monolingual corpus disambiguation by the approaches of Lesk : University of Montreal, Faculty of Arts and Sciences; Paper presented at the Faculty of Graduate Studies to obtain the rank of Master of Science (MSc) in computer science. [find](https://scholar.google.com/scholar?q=Florentina+Vasilescu+%2C+2003%2E+Monolingual+corpus+disambiguation+by+the+approaches+of+Lesk+%3A+University+of+Montreal%2C+Faculty+of+Arts+and+Sciences%3B+Paper+presented+at+the+Faculty+of+Graduate+Studies+to+o)
8. Harman D., 1986. An experimental study of factors important in document ranking ; Actes de ACM Conference on Research and Development in Information Retrieval ; Pise, Italy ; 1986. [find](https://scholar.google.com/scholar?q=Harman+D%2E%2C+1986%2E+An+experimental+study+of+factors+important+in+document+ranking+%3B+Actes+de+ACM+Conference+on+Research+and+Development+in+Information+Retrieval+%3B+Pise%2C+Italy+%3B+1986%2E)
9. Karov, Yael and Edelman, Shimon (1998). “Similarity-based word sense disambiguation”. In this issue. [find](https://scholar.google.com/scholar?q=Karov%2C+Yael+and+Edelman%2C+Shimon+%281998%29%2E+%E2%80%9CSimilarity%2Dbased+word+sense+disambiguation%E2%80%9D%2E+In+this+issue%2E)
10. Larkey L. S., Ballesteros L. and Connell M., « Improving Stemming for Arabic Information Retrieval : Light Stemming and Cooccurrence Analysis », In Proceedings of the 25 th Annual International Conference on Research and Development in Information Retrieval (SIGIR 2002), Tampere, Finland, August 2002, p. 275-282 [find](https://scholar.google.com/scholar?q=Larkey+L%2E+S%2E%2C+Ballesteros+L%2E+and+Connell+M%2E%2C+%C2%AB+Improving+Stemming+for+Arabic+Information+Retrieval+%3A+Light+Stemming+and+Cooccurrence+Analysis+%C2%BB%2C+In+Proceedings+of+the+25+th+Annual+International+Confer)
11. Michael Lesk, Automatic sense disambiguation using machine readable dictionaries: how to tell a pine cone from an ice cream cone , ACM Special Interest Group for Design of Communication Proceedings of the 5th annual international conference on Systems documentation, p. 24 - 26, 1986. ISBN 0897912241 [find](https://scholar.google.com/scholar?q=Michael+Lesk%2C+Automatic+sense+disambiguation+using+machine+readable+dictionaries%3A+how+to+tell+a+pine+cone+from+an+ice+cream+cone+%2C+ACM+Special+Interest+Group+for+Design+of+Communication+Proceedings+of)
12. Nancy Ide, Jean Verronis. 1998.Word Sense Disambiguation: The State Of The Art. Computational Linguistics, 2424:1, 1-40. [find](https://scholar.google.com/scholar?q=Nancy+Ide%2C+Jean+Verronis%2E+1998%2EWord+Sense+Disambiguation%3A+The+State+Of+The+Art%2E+Computational+Linguistics%2C+2424%3A1%2C+1%2D40%2E)
13. Robertson et al., 1994.: S. Robertson, S. Walker, M. Hancock-Beaulieu, M. Gatford ; [Okapi at TREC-3](https://md.research.elamur.ai/free/en/papers/robertson-1994-okapi-trec-3/) ; Acts de Third Text Retrieval Conference (TREC-3), NIST special publication 500-225 ; pp. 109-126 ; Gaithersburg, Maryland, USA ; 1994.
14. Salton, G. & Buckley, C. (1988). Term-weighting approaches in automatic text retrieval. Information Processing and Management, 24(5), 513-523. [↗](https://doi.org/10.1016/0306-4573(88)90021-0)
15. Zouaghi A., Zrigui M., Antoniadis G. 2008. Understanding of the Arabic spontaneous speech: A numeric modelisation, Revue TAL VARIA. [find](https://scholar.google.com/scholar?q=Zouaghi+A%2E%2C+Zrigui+M%2E%2C+Antoniadis+G%2E+2008%2E+Understanding+of+the+Arabic+spontaneous+speech%3A+A+numeric+modelisation%2C+Revue+TAL+VARIA%2E)

*Updated 2 November 2026*