Elamur Research⌕
☰
home / papers / 2011 / Developing a New System for Arabic Morphological Analysis and Generation

Developing a New System for Arabic Morphological Analysis and Generation

Paper · 2011 · Mourad Gridach and Noureddine Chenfour

Summary

The paper addresses computational morphology for Modern Standard Arabic. Arabic builds words by applying schemes (patterns) to consonantal roots and then attaching prefixes and suffixes, which makes analysis hard. The authors first encode an Arabic morphological database in XMODEL, an XML-based Morphological Definition Language. The database is organised into morphological classes, properties (feature descriptors such as gender) and rules that combine components into valid words. From this database they compile a family of morphological automata (AMAUT), one per morphological category. A framework then runs a pipeline over input text: tokenize, select the suitable automata, determine part of speech, find possible affixes, and output rich morpho-syntactic features. The same resource serves both analysis and generation. On an ALECSO evaluation corpus, the system is compared with ElixirFM and the Xerox analyzer, and the authors report the highest word-level accuracy. The work matters because the database stays compact and independent of processing, while the output carries more features than competing tools.

Key points

  • The XMODEL lexicon has 5970 entries, and the compiled set of AMAUT automata takes about 120 MB.
  • The analysis pipeline has five steps: tokenization, loading the automata, POS detection, affix extraction, and feature extraction based on the valid affixes and the scheme.
  • Verbs receive several features: type (strong, weak, incomplete), tense (mADI, muDAric, eamr) and voice. Nouns are marked as derived or particular, and particles are subtyped, for example preposition or conjunction.
  • Evaluation used text from ALECSO, which ran an Arabic morphological analyzer competition in Damascus in April 2009. Accuracy was defined as the share of words that received a good solution.
  • Reported accuracies are about 88.91% and 89.58% for the Xerox analyzer and ElixirFM (the extracted table is garbled) and 95.08% for the proposed system. For contrast, the paper describes Xerox as having 4930 roots and 400 patterns giving about 90,000 stems, and Buckwalter as having 77,800 stems.

Interesting ideas

  • I1. The lexicon stores reusable classes, feature properties and combination rules instead of listing every surface form, which keeps the resource small and independent of the processing code.
  • I2. A single declarative resource drives both analysis and generation, unlike analyzers that cannot easily be run in reverse.
  • I3. Morphology is split into many category-specific automata, and only those that fit a given word are activated.
  • I4. The authors criticise a rule-based competitor for overgenerating forms that no dictionary attests, which shows that a generative rule system needs a check against attested words.
  • I5. The evaluation scores a word as correct when any good solution appears among its outputs, and reports results per POS group (nouns, verbs, particles).

For Mova

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

  • I1 (tokens): The en morphology could keep inflection classes and affix rules declaratively in data files and compile them, so lexicon growth does not require rewriting the analyzer.
  • I2 (back): Realising text from the meaning graph could use the same morphology tables as the tokenizer, so a lemma plus features produced in generation is guaranteed to be analysable again.
  • I3 (tokens): Separate small finite-state components for verbs, nouns and closed-class words could each be inspected on their own, which makes every lemma decision traceable to one component.
  • I4 (learn): Self-induced morphology or repair rules could be tested against a list of attested forms as a negative control, flagging rules that invent non-words.
  • I5 (learn): Morphology evaluation in en could report per-POS coverage and "correct analysis among candidates" separately from top-1 accuracy, which shows whether errors come from missing analyses or from bad ranking.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 2Research subjects and languages the work is about.Semitic languages · 2,908 worksMorphology · 405 works

Cite

Cite as Mourad Gridach and Noureddine Chenfour. 2011. Developing a New System for Arabic Morphological Analysis and Generation. Proceedings of the 2nd Workshop on South Southeast Asian Natural Language Processing (WSSANLP).

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

@inproceedings{gridach2011developing,
  title = {{Developing a New System for Arabic Morphological Analysis and Generation}},
  author = {Mourad Gridach and Noureddine Chenfour},
  booktitle = {Proceedings of the 2nd Workshop on South Southeast Asian Natural Language Processing (WSSANLP)},
  year = {2011},
  url = {https://aclanthology.org/W11-3009/}
}

References

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

  1. Al-Sughaiyer Imad A. and Al-Kharashi Ibrahim A. 2004. Arabic morphological analysis techniques: A comprehensive survey. Journal of the American Society for Information Science and Technology, 55(3):189–213 ↗
  2. Altantawy Mohamed, Nizar Habash, Owen Rambow, and Ibrahim Saleh. 2010. Morphological Analysis and Generation of Arabic Nouns: A Morphemic Functional Approach. In Proceedings of the Language Resource and Evaluation Conference (LREC-2010), Malta.
  3. Attia, M. 2006. An Ambiguity-Controlled Morphological Analyzer for Modern Standard Arabic Modelling Finite State Networks. The Challenge of Arabic for NLP/MT Conference, the British Computer Society, London. find
  4. Atwell E., Al-Sulaiti L., Al-Osaimi S., Abu Shawar B.. 2004. A Review of Arabic Corpus Analysis Tools, JEP-TALN 04, Arabic Language Processing, Fès, 19-22 April. find
  5. Beesley KR 1996. Arabic Finite-State Morphological Analysis and Generation, Proceedings of the 16th conference on Computational linguistics, Vol 1. Copenhagen, Denmark: Association for Computational Linguistics, pp 89-94.
  6. Beesley KR. 2000. Finite-State Non-Concatenative Morphotactics SIGPHON-2000, Proceedings of the Fifth Workshop of the ACL Special Interest Group in Computational Phonology, p. 1-12, August 6, 2000, Luxembourg. find
  7. Buckwalter T. 2002. Buckwalter Arabic Morphological Analyzer Version 1.0. Linguistic Data Consortium, University of Pennsylvania, LDC Catalog No.: LDC2002L49. find
  8. Buckwalter, T. 2004. Buckwalter Arabic Morphological Analyzer Version 2.0. Linguistic Data Consortium, catalog number LDC2004L02 and ISBN 1-58563-324-0. Cavalli-Sforza, V., Soudi. find
  9. A, and Teruko M. 2000. Arabic Morphology Generation Using a Concatenative Strategy. In Proceedings of the 1st Conference of the North American Chapter of the Association for Computational Linguistics (NAACL 2000), Seattle, USA.
  10. Darwish K. (2002). Building a Shallow Morphological Analyzer in One Day, Proceedings of the workshop on Computational Approaches to Semitic Languages in the 40th Annual Meeting of the Association for Computational Linguistics (ACL-02). Philadelphia, PA, USA. find
  11. Habash Nizar. 2004. Large scale lexeme based Arabic orphological generation. In Proceedings of Traitement Automatique du Langage Naturel (TALN-04). Fez, Morocco. find
  12. Habash Nizar and Rambow Owen. 2005. Arabic Tokenization, Part-of-Speech Tagging and Morphological Disambiguation in One Fell Swoop. In Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics ACL, pages 573–580, Ann Arbor.
  13. Forsberg M. and Ranta A. 2004. Functional Morphology ICFP'04, Proceedings of the Ninth ACM SIGPLAN International Conference of Functional Programming, September 19-21, Snowbird, Utah Ibrahim, K. 2002. Al-Murshid fi Qawa'id Al-Nahw wa Al-Sarf [The Guide in Syntax and Morphology Rules]. Amman, Jordan, Al-Ahliyyah for Publishing and Distribution. find
  14. Otakar Smrz (2007). ElixirFM. Implementation of Functional Arabic Morphology. In ACL Proceedings of the Workshop on Computational Approaches to Semitic Languages: Common Issues and Resources, pages 1–8, Prague, Czech Republic.
  15. Otakar Smrž and Viktor Bielický. 2010. ElixirFM. Functional Arabic Morphology, http://sourceforge.net/projects/elixir-fm/. ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/gridach-2011-developing-new-system-arabic-morphological/