---
title: "A Multi-Word Term Extraction Program for Arabic Language (2008)"
description: "Because Arabic is agglutinative and morphologically rich, purely statistical counting does not work… 22 references, cited by 13 papers here."
authors: ["Siham Boulaknadel", "Beatrice Daille", "Driss Aboutajdine"]
year: 2008
venue: "Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08)"
source: "https://aclanthology.org/L08-1155/"
licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)"
licence_url: "https://creativecommons.org/licenses/by-nc/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/boulaknadel-2008-multi-word-term-extraction-program/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/boulaknadel-2008-multi-word-term-extraction-program/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2008](https://md.research.elamur.ai/free/en/papers/2008/) / A Multi-Word Term Extraction Program for Arabic Language

# A Multi-Word Term Extraction Program for Arabic Language

**Authors:** Siham Boulaknadel, [Beatrice Daille](https://md.research.elamur.ai/free/en/authors/beatrice-daille/) and Driss Aboutajdine · **Year:** 2008 · **Venue:** [Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08)](https://md.research.elamur.ai/free/en/venues/lrec/) ([LREC 2008](https://md.research.elamur.ai/free/en/venues/lrec/2008/)) · **Publisher:** [European Language Resources Association (ELRA)](https://md.research.elamur.ai/free/en/publishers/european-language-resources-association-elra/) · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/L08-1155/) · [PDF](https://aclanthology.org/L08-1155.pdf) · [OpenAlex](https://openalex.org/W96139068)

**Citations:** 42 citations (OpenAlex, 2026-10-14) · [13](https://md.research.elamur.ai/free/en/papers/boulaknadel-2008-multi-word-term-extraction-program/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The paper tackles automatic terminology extraction for Arabic. Because Arabic is agglutinative and morphologically rich, purely statistical counting does not work well, so the authors build a hybrid pipeline. Candidate multi-word terms (MWTs) are first selected with part-of-speech patterns over output from Diab's tagger. The pattern grammar targets two-content-word structures such as N ADJ, N1 N2 and N1 PREP N2, and it accounts for graphical, inflectional, morphosyntactic and syntactic variation. The candidates are then ranked with four association measures: LLR, FLR, MI3 and t-score. Since no specialised Arabic corpus existed, the authors compiled one from two environmental news websites. They also assembled a reference term list drawing on resources such as Agrovoc and used it to judge the rankings automatically. LLR gives the best precision and MI3 the worst, which mirrors earlier findings for Romance languages. The work fills a gap in tools for specialised Arabic language processing and could support information retrieval and information extraction.

## Key points
- The environment-domain corpus has 1,062 documents and 475,148 tokens, of which 54,705 are distinct. It was collected from the "Al-Khat Alakhdar" and "Akhbar Albiae" websites.
- The reference list contains about 65,000 Arabic environmental MWTs gathered from known terminology sources such as Agrovoc.
- Precision by measure: LLR 85%, FLR 60%, t-score 57%, MI3 26%. The hypothesis-testing measures clearly beat the information-theoretic MI3.
- The candidate grammar maps the patterns N1 N2, N1 ADJ and N1 PREP N2 (with sub-patterns using the prepositions b, l, mn) onto tag sequences from Diab's tagger. Arabic examples are written in Buckwalter transliteration.
- The variant typology follows Daille (2005):
  - graphical variants (the p/h letter alternation);
  - inflectional variants (number, gender, the definite prefix Al);
  - morphosyntactic synonymy (N ADJ ⇔ N PREP N);
  - syntactic variants (insertion, postposition, coordination).

## Interesting ideas
- Treat structurally different phrasings as the same term, for example an adjective form and a prepositional form ("oil wells" vs. "wells of oil"), instead of counting them as separate units.
- Use an explicit typology of term variation (insertion, postposition, coordination) to recognise a base term inside an expanded or coordinated phrase.
- Filter candidates with linguistic patterns first and rank them statistically only afterwards, so the counts never see ungrammatical n-grams.
- Evaluate rankings automatically against an existing external term base rather than by hand judgement. This makes it cheap to compare several scoring functions on the same footing.
- Normalise orthographic alternations, such as the Arabic p/h letter swap, before counting, so that spelling noise does not fragment frequencies.

## For Mova

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

- **I1** (world): Entity resolution could merge "chemical pollution" and "pollution by chemicals" into one concept node. That would keep the world of the text from splitting one thing into two.
- **I2** (tree): After UD parsing, rules over amod/nmod/conj edges could recover base compounds from "pollution of soil and water". The two distributed terms would then become explainable.
- **I3** (world): Compiled knowledge (measures, ontology links) could be mined from corpora by ranking only tree-pattern-validated noun compounds. This keeps the result explainable and avoids bag-of-words noise.
- **I4** (learn): Silver term or frame lexicons could be scored against WordNet/Wikidata entries as an automatic yardstick. Comparing several rankers side by side would expose weak scorers, much as MI3 was exposed here.
- **I5** (tokens): The English lemmatizer could canonicalise spelling variants (British/American, hyphenated vs. joined compounds) before frequency-based decisions.

*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

- [Centre National de la Recherche Scientifique](https://md.research.elamur.ai/free/en/labs/centre-national-de-la-recherche-scientifique/) — Research institute, Paris, France · 168 works · 62,369 citations
- [Nantes Université](https://md.research.elamur.ai/free/en/universities/nantes-universite/) — University, Nantes, France · 5 works · 470 citations
- [Université de Nantes](https://md.research.elamur.ai/free/en/universities/universite-de-nantes/) — University, Nantes, France · 2 works · 61 citations
- [Laboratoire d'informatique de Nantes Atlantique](https://md.research.elamur.ai/free/en/labs/laboratoire-d-informatique-de-nantes-atlantique/) — Research institute, Nantes, France · 2 works · 55 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 · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (1): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 2,908 works

## Cite

Cite as: Siham Boulaknadel, Beatrice Daille and Driss Aboutajdine. 2008. A Multi-Word Term Extraction Program for Arabic Language. Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08).

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

```bibtex
@inproceedings{boulaknadel2008multi,
  title = {{A Multi-Word Term Extraction Program for Arabic Language}},
  author = {Siham Boulaknadel, Beatrice Daille and Driss Aboutajdine},
  booktitle = {Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08)},
  year = {2008},
  doi = {10.63317/4eafyvur9wua},
  url = {https://doi.org/10.63317/4eafyvur9wua}
}
```

- [← previous Improving NER in Arabic Using a Morphological Tagger](https://md.research.elamur.ai/free/en/papers/farber-2008-improving-ner-arabic-morphological-tagger/)
- [next → Contradictions and Justifications: Extensions to the Textual Entailment Task](https://md.research.elamur.ai/free/en/papers/voorhees-2008-contradictions-justifications-extensions-textual-entailment/)
- [↑ back to papers of 2008](https://md.research.elamur.ai/free/en/papers/2008/)

## References

22 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. P. Ahlgren and J. Keklinen. 2006. Swedish full text retrieval: Effectiveness of different combinations of indexing strategies with query terms. Information Retrieval, 9(6):681–697. [↗](https://doi.org/10.1007/s10791-006-9009-1)
2. M.T. Cabré, R. ESTOP, and J. VIVALDI. 2000. Automatic term detection: a review of current systems. Recent Advances in Computational Terminology, 2(1):53–88. [find](https://scholar.google.com/scholar?q=M%2ET%2E+Cabre%CC%81%2C+R%2E+ESTOP%2C+and+J%2E+VIVALDI%2E+2000%2E+Automatic+term+detection%3A+a+review+of+current+systems%2E+Recent+Advances+in+Computational+Terminology%2C+2%281%29%3A53%E2%80%9388%2E)
3. K.W. Church and P. Hanks. 1990. [Word association norms, mutual information, and lexicography](https://md.research.elamur.ai/free/en/papers/church-1990-word-association-norms-mutual-information/). In Proceedings of the 27th. Annual Meeting of the Association for Computational Linguistics, pages 76–83, Vancouver, B.C. Association for Computational Linguistics. [↗](https://doi.org/10.5555/89086.89095)
4. K. Church, W. Gale, P. Hanks, and D. Hindle. 1991. Using statistics in lexical analysis. In Lexical Acquisition: Exploiting On-Line Resources to Build a Lexicon, pages 115–164. U. Zernik. [find](https://scholar.google.com/scholar?q=K%2E+Church%2C+W%2E+Gale%2C+P%2E+Hanks%2C+and+D%2E+Hindle%2E+1991%2E+Using+statistics+in+lexical+analysis%2E+In+Lexical+Acquisition%3A+Exploiting+On%2DLine+Resources+to+Build+a+Lexicon%2C+pages+115%E2%80%93164%2E+U%2E+Zernik%2E)
5. B. Daille. 1994. Approche mixte pour l’extraction de terminologie : statistiques lexicales et filtres linguistiques. Ph.D. thesis, Universit de Paris 7, France. [find](https://scholar.google.com/scholar?q=B%2E+Daille%2E+1994%2E+Approche+mixte+pour+l%E2%80%99extraction+de+terminologie+%3A+statistiques+lexicales+et+filtres+linguistiques%2E+Ph%2ED%2E+thesis%2C+Universit+de+Paris+7%2C+France%2E)
6. B. Daille. 2005. Variations and application-oriented terminology engineering. International journal of theoretical and applied issues in specialized communication, 11(1):181–197. [↗](https://doi.org/10.1075/term.11.1.08dai)
7. M. Diab, K. Hacioglu, and D. Jurafsky. 2004. [Automatic tagging of arabic text: From raw text to base phrase chunks](https://md.research.elamur.ai/free/en/papers/diab-2004-automatic-tagging-arabic-text-raw/). In In Proceedings of NAACL-HLT, pages 149– 152, Boston, USA.
8. T. Dunning. 1994. [Accurate methods for the statistics of surprise and coincidence](https://md.research.elamur.ai/free/en/papers/dunning-1993-accurate-methods-statistics-surprise-coincidence/). Computational Linguistics, 19(1):61–74.
9. H. Haddad. 2002. Extraction et impact des connaissances sur les performances des systmes de recherche 1488 d’information. Ph.D. thesis, Universit Joseph Fourier, Grenoble, France. [find](https://scholar.google.com/scholar?q=H%2E+Haddad%2E+2002%2E+Extraction+et+impact+des+connaissances+sur+les+performances+des+systmes+de+recherche+1488+d%E2%80%99information%2E+Ph%2ED%2E+thesis%2C+Universit+Joseph+Fourier%2C+Grenoble%2C+France%2E)
10. M. Hong, S. Fissaha, and J. Haller. 2001. Hybrid filtering for extraction of term candidates from german technical texts. In TIA 2001 : terminologie et intelligence artificielle, pages 223–232. [find](https://scholar.google.com/scholar?q=M%2E+Hong%2C+S%2E+Fissaha%2C+and+J%2E+Haller%2E+2001%2E+Hybrid+filtering+for+extraction+of+term+candidates+from+german+technical+texts%2E+In+TIA+2001+%3A+terminologie+et+intelligence+artificielle%2C+pages+223%E2%80%93232%2E)
11. F. Ibekwe-SanJuan and M. T. Condamines, A.and Cabr Castellv. 2007. Application-driven terminology engineering. special issue of the Terminology Journal, 2:1–17. [find](https://scholar.google.com/scholar?q=F%2E+Ibekwe%2DSanJuan+and+M%2E+T%2E+Condamines%2C+A%2Eand+Cabr+Castellv%2E+2007%2E+Application%2Ddriven+terminology+engineering%2E+special+issue+of+the+Terminology+Journal%2C+2%3A1%E2%80%9317%2E)
12. C. Jacquemin. 2001. Spotting and Discovering Terms through Natural Language Processing Techniques. MIT Press, Cambridge. [find](https://scholar.google.com/scholar?q=C%2E+Jacquemin%2E+2001%2E+Spotting+and+Discovering+Terms+through+Natural+Language+Processing+Techniques%2E+MIT+Press%2C+Cambridge%2E)
13. W. C Kenneth and P. Hanks. 1989. [Word association norms, mutual information, and lexicography](https://md.research.elamur.ai/free/en/papers/church-1990-word-association-norms-mutual-information/). In Proceedings of the 27th. Annual Meeting of the Association for Computational Linguistics, pages 76–83, Vancouver, B.C.
14. R.H. Marcelline, K.S. Guergana, M.J. Thomas, and Christopher G.C. 2003. A term extraction tool for expanding content in the domain of functioning, disability, and health: proof of concept. Journal of Biomedical Informatics, 36(4/5):250–259. [find](https://scholar.google.com/scholar?q=R%2EH%2E+Marcelline%2C+K%2ES%2E+Guergana%2C+M%2EJ%2E+Thomas%2C+and+Christopher+G%2EC%2E+2003%2E+A+term+extraction+tool+for+expanding+content+in+the+domain+of+functioning%2C+disability%2C+and+health%3A+proof+of+concept%2E+Journal+of+)
15. G. Maynard and S. Ananiadou. 2000. Identifying terms by their family and friends. In Proceedings of the 18th conference on Computational linguistics, pages 530 – 536, Saarbrcken, Germany. [find](https://scholar.google.com/scholar?q=G%2E+Maynard+and+S%2E+Ananiadou%2E+2000%2E+Identifying+terms+by+their+family+and+friends%2E+In+Proceedings+of+the+18th+conference+on+Computational+linguistics%2C+pages+530+%E2%80%93+536%2C+Saarbrcken%2C+Germany%2E)
16. A. Michiels and N. Dufour. 1998. Defi, a tool for automatic multi-word unit rec-ognition, meaning assignment and translation selection. In In Proceedings of the First Interna-tional Conference on Language Resources and Evaluation, pages 1179–1186, Granada, Spain. [find](https://scholar.google.com/scholar?q=A%2E+Michiels+and+N%2E+Dufour%2E+1998%2E+Defi%2C+a+tool+for+automatic+multi%2Dword+unit+rec%2Dognition%2C+meaning+assignment+and+translation+selection%2E+In+In+Proceedings+of+the+First+Interna%2Dtional+Conference+on+Lang)
17. H. Nakagawa and T. Mori. 2002. Nested collocation and compound noun for term recognition. In In Proceedings of the First Workshop on Computational Terminology COMPTERM’98,, pages 64–70. [find](https://scholar.google.com/scholar?q=H%2E+Nakagawa+and+T%2E+Mori%2E+2002%2E+Nested+collocation+and+compound+noun+for+term+recognition%2E+In+In+Proceedings+of+the+First+Workshop+on+Computational+Terminology+COMPTERM%E2%80%9998%2C%2C+pages+64%E2%80%9370%2E)
18. H. Nakagawa and T. Mori. 2003. Automatic term recognition based on statistics of compound nouns and their components. Terminology, 9(2):201–219. [↗](https://doi.org/10.1075/term.9.2.04nak)
19. G. Nenadic and S. Spasic, I.and Ananiadou. 2002. Automatic acronym acquisition and term variation management within domain-specific texts. In In: Proceedings of 3rd International Conference on Language, Resources and Evaluation, LREC-3, pages 2155–2162, Las Palmas, Spain. [find](https://scholar.google.com/scholar?q=G%2E+Nenadic+and+S%2E+Spasic%2C+I%2Eand+Ananiadou%2E+2002%2E+Automatic+acronym+acquisition+and+term+variation+management+within+domain%2Dspecific+texts%2E+In+In%3A+Proceedings+of+3rd+International+Conference+on+Languag)
20. A. Roman. 1990. La Grammaire de l’arabe. PUF, Paris. [find](https://scholar.google.com/scholar?q=A%2E+Roman%2E+1990%2E+La+Grammaire+de+l%E2%80%99arabe%2E+PUF%2C+Paris%2E)
21. I.A. Sag, F. Baldwin, T.and Bond, A. Copestake, and D. Flickinger. 2002. [Multiword expressions: A pain in the neck for NLP](https://md.research.elamur.ai/free/en/papers/sag-2002-multiword-expressions-pain-neck-nlp/). In CICLing, pages 1–15.
22. F. Smadja. 1993. Xtract : An overview. Computers and the Humanities, 26:399–413. [↗](https://doi.org/10.1007/bf00136983)

*Updated 1 November 2026*