Elamur Research⌕
☰
home / papers / 2017 / Morphology-based Entity and Relational Entity Extraction Framework for Arabic

Morphology-based Entity and Relational Entity Extraction Framework for Arabic

Paper · 2017 · Amin Jaber and Fadi A. Zaraket

Summary

Rule-based tools for extracting entities and relations usually require programming and linguistic expertise, and they rarely handle Arabic morphology. This is a serious gap because Arabic omits short vowels, has rich affixation, and can even make word boundaries ambiguous. The authors present MERF, an open-source framework in which users build extraction rules interactively. The building blocks are Boolean formulae over morphological features from the Sarf analyzer, which is based on finite state transducers. These features include prefix, stem, suffix, POS, gloss, synonym and semantic-category tags. Users combine the formulae into regular expressions, and MERF compiles each expression to a non-deterministic automaton and runs it over the formula matches. Users can attach C++ actions to subexpressions and declare relations as labelled source–destination tuples, from which MERF builds an entity graph. In several case studies, MERF needed less development effort than application-specific tools and gave reasonably accurate results at acceptable runtime cost. It thus offers a practical way to create Arabic corpora annotated with entities and relations.

Key points

  • MERF builds on Sarf, an in-house Arabic morphological analyzer that returns every possible morphological solution for a word. Each solution is a set of morphemes carrying POS, gloss and category tags.
  • The regex operators are:
    • concatenation;
    • optional (zero or one);
    • Kleene star and plus;
    • bounded repetition up to M;
    • conjunction and disjunction.
  • Named expressions together form a local-grammar-style rule set.
  • User actions are written in C++ against an API that exposes a match's text, position, length, numeric value and morphology. MERF generates the code, compiles it and loads it as shared libraries.
  • A "light Arabic WordNet" relation uses the synonymity feature Syn k. MERF extends the earlier MATAr tagger (Jaber and Zaraket, 2013) and is released at github.com/codelogicanalysis/atmine.
  • The running example is driving directions to Dubai Mall. Its extracted entities are persons, places, relative positions and numbers, which are linked into a graph of relations.

Interesting ideas

  • I1. Pattern atoms are Boolean tests over all candidate morphological analyses of a word rather than over surface strings, so ambiguity survives until a rule settles it
  • I2. Relations are declared as (source, destination, label) tuples that point to named capture groups of a rule, so a single match directly produces graph edges
  • I3. A synonym-expansion feature (Syn k, a light WordNet relation) can be used as an ordinary matchable term, so one rule covers a whole family of near-synonyms
  • I4. Rules can carry small attached actions that compute values from a match, such as numeric value, position or length
  • I5. A built-in viewer compares automatic matches against reference tags to estimate accuracy while rules are being written

For Mova

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

  • I1 (tokens): Rules in crate en could match over the full set of lemma/feature alternatives for a token, and record which reading each rule chose, so that disambiguation remains explainable.
  • I2 (world): World-building rules could emit typed edges between entities with the ID of the rule that produced them, which gives every fact in the text world a traceable origin.
  • I3 (frames): Frame triggers could match WordNet synonym sets instead of single lemmas, widening VerbNet role assignment without falling back to bag-of-words similarity.
  • I4 (answer): Maths word-problem readers could attach quantity-normalising actions to number and unit patterns, so extracted measures arrive already typed and valued.
  • I5 (learn): Self-induced repair rules could be scored continuously against gold or silver tags in the same way, flagging a rule as soon as it starts hurting precision.

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 Amin Jaber and Fadi A. Zaraket. 2017. Morphology-based Entity and Relational Entity Extraction Framework for Arabic. arXiv (Cornell University).

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

@misc{jaber2017morphology,
  title = {{Morphology-based Entity and Relational Entity Extraction Framework for Arabic}},
  author = {Amin Jaber and Fadi A. Zaraket},
  year = {2017},
  url = {https://aclanthology.org/2017.tal-3.6/}
}

References

45 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. AbdelRahman S., Elarnaoty M., Magdy M., Fahmy A., “Integrated Machine Learning Techniques for Arabic Named Entity Recognition”, International Journal of Computer Science Issues, vol. 7, no 4, p. 27-36, 2010. find
  2. Al-Sughaiyer I., Al-Kharashi I., “Arabic morphological analysis techniques: A comprehensive survey”, JASIST, 2003. ↗
  3. Alrahabi M., Ibrahim A. H., Desclés J.-P., “Semantic Annotation of Reported Information in Arabic”, FLAIRS Conference, vol. 6, p. 263-268, 2006. find
  4. Appelt D., Onyshkevych B., “The common pattern specification language”, TIPSTER workshop, ACL, 1998. ↗
  5. Attia M., Rashwan M., Al-Badrashiny M., “Fassieh, a semi-automatic visual interactive tool for morphological, PoS-Tags, phonetic, and semantic annotation of Arabic text corpora”, IEEE transactions on audio, speech, and language processing, vol. 17, no 5, p. 916-925, 2009. ↗
  6. Atzmueller M., Kluegl P., Puppe F., “Rule-Based Information Extraction for Structured Data Acquisition using TextMarker”, Proceedings of LWA, Citeseer, p. 1-7, 2008. find
  7. Benajiba Y., Diab M., Rosso P., “Arabic named entity recognition using optimized feature sets”, Proceedings of the Conference on Empirical Methods in Natural Language Processing, Association for Computational Linguistics, p. 284-293, 2008.
  8. Benajiba Y., Rosso P., Benedíruiz J. M., “Anersys: An Arabic named entity recognition system based on maximum entropy”, International Conference on Intelligent Text Processing and Computational Linguistics, Springer, p. 143-153, 2007. find
  9. Buckwalter T., Buckwalter Arabic Morphological Analyzer Version 1.0, Technical report, University of Pennsylvania, 2002. find
  10. Chiticariu L., Krishnamurthy R., Li Y., Raghavan S., Reiss F. R., Vaithyanathan S., “SystemT: an algebraic approach to declarative information extraction”, Proceedings of the Association for Computational Linguistics, p. 128-137, 2010. find
  11. Dorr B. J., Passonneau R. J., Farwell D., Green R., Habash N., Helmreich S., Hovy E., Levin L., Miller K. J., Mitamura T. et al., “Interlingual annotation of parallel text corpora: a new framework for annotation and evaluation”, NLE, vol. 16, no 3, p. 197-243, 2010. ↗
  12. Dukes K., Atwell E., Habash N., “Supervised collaboration for syntactic annotation of Quranic Arabic”, Language resources and evaluation, vol. 47, no 1, p. 33-62, 2013. find
  13. Ekbal A., Bandyopadhyay S., “Named Entity Recognition using Support Vector Machine: A Language Independent Approach”, IJCSSE, 2008. find
  14. Ferilli S., “Natural Language Processing”, Automatic Digital Document Processing and Management, Springer, 2011. ↗
  15. Gennari J., Musen M., Fergerson R., Grosso W., Crubézy M., Eriksson H., Noy N., Tu S., “The evolution of Protégé: an environment for knowledge-based systems development”, International Journal of Human-Computer Studies, vol. 58, no 1, p. 89-123, 2003. ↗
  16. Habash N., “Introduction to Arabic natural language processing”, Synthesis Lectures on Human Language Technologies, 2010. ↗
  17. Habash N., Sadat F., “Arabic Preprocessing Schemes for Statistical Machine Translation”, NAACL, p. 49-52, 2006.
  18. Hammo B., Abu-Salem H., Lytinen S., “QARAB: A question answering system to support the Arabic language”, Computational approaches to semitic languages, ACL, p. 1-11, 2002. — QARAB: A: Question Answering System to Support the Arabic Language
  19. Jaber A., Zaraket F., “MATAr: Morphology-based Tagger for Arabic”, AICCSA, May, 2013. find
  20. Kholidy H., Chatterjee N., “Towards developing an Arabic word alignment annotation tool with some Arabic alignment guidelines”, ISDA, IEEE, p. 778-783, 2010. find
  21. Kulick S., “Consistent and flexible integration of morphological annotation in the Arabic Treebank”, LREC, 2010. find
  22. Linckels S., Meinel C., “Natural Language Processing”, E-Librarian Service, Springer, p. 61-79, 2011. ↗
  23. Maamouri M., Bies A., Buckwalter T., Mekki W., “The penn Arabic treebank: Building a large-scale annotated Arabic corpus”, NEMLAR Conference on Arabic Language Resources and Tools, p. 102-109, 2004. find
  24. MERF Makhlouta J., Zaraket F., Harkous H., “Arabic entity graph extraction using morphology, finite state machines, and graph transformations”, CICLing, Springer, p. 297-310, 2012. find
  25. Marcus M. P., Marcinkiewicz M. A., Santorini B., “Building a large annotated corpus of English: The Penn Treebank”, Computational linguistics, vol. 19, no 2, p. 313-330, 1993. ↗
  26. Miller G. A., Beckwith R., Fellbaum C., Gross D., Miller K. J., “Introduction to WordNet: An on-line lexical database”, International journal of lexicography, vol. 3, no 4, p. 235-244, 1990. ↗
  27. Morton T., LaCivita J., “WordFreak: an open tool for linguistic annotation”, HLT/NAACL, 2003. find
  28. Müller C., Strube M., “Multi-level annotation of linguistic data with MMAX2”, Corpus technology and language pedagogy: New resources, new tools, new methods, 2006. find
  29. Nolan D., Lang D., “JavaScript Object Notation”, XML and Web Technologies for Data Sciences with R, Springer, 2014. ↗
  30. Ogren P., “Knowtator: a protégé plug-in for annotated corpus construction”, NAACL-Demonstrations, ACL, 2006. find
  31. Pasha A., Al-Badrashiny M., Diab M. T., El Kholy A., Eskander R., Habash N., Pooleery M., Rambow O., Roth R., “MADAMIRA: A Fast, Comprehensive Tool for Morphological Analysis and Disambiguation of Arabic”, LREC, vol. 14, p. 1094-1101, 2014.
  32. Saleh L., Al-Khalifa H., “AraTation: an Arabic semantic annotation tool”, IIWAS, 2009. find
  33. Settles B., “Closing the loop: Fast, interactive semi-supervised annotation with queries on features and instances”, Proceedings of EMNLP, ACL, p. 1467-1478, 2011. find
  34. Shahrour A., Khalifa S., Taji D., Habash N., “Camelparser: A system for Arabic syntactic analysis and morphological disambiguation”, COLING Demonstrations, p. 228-232, 2016.
  35. Sipser M., Introduction to the Theory of Computation, Cengage Learning, 2012. find
  36. Smrz O., Pajas P., “Morphotrees of Arabic and their annotation in the TrEd environment”, NEMLAR International Conference on Arabic Language Resources and Tools, 2004. find
  37. Soudi A., Neumann G., Van den Bosch A., Arabic computational morphology: knowledge-based and empirical methods, Springer, 2007. find
  38. Stenetorp P., Pyysalo S., Topić G., Ohta T., Ananiadou S., Tsujii J., “BRAT: a web-based tool for NLP-assisted text annotation”, EACL Demonstrations, ACL, p. 102-107, 2012.
  39. Traboulsi H., “Arabic named entity extraction: A local grammar-based approach”, IMCSIT, IEEE, 2009. find
  40. Urbain J., “User-driven relational models for entity-relation search and extraction”, Proceedings of JIWES, ACM, 2012. find
  41. Xue N., Xia F., Chiou F.-D., Palmer M., “The Penn Chinese TreeBank: Phrase structure annotation of a large corpus”, Natural language engineering, vol. 11, no 2, p. 207-238, 2005. ↗
  42. Zaghouani W., Pouliquen B., Ebrahim M., Steinberger R., “Adapting a resource-light highly multilingual Named Entity Recognition system to Arabic”, LREC, p. 563-567, 2010.
  43. Zaraket F. A., Makhlouta J., “Arabic Cross-Document NLP for the Hadith and Biography Literature”, FLAIRS, May, 2012a. find
  44. Zaraket F., Makhlouta J., “Arabic Morphological Analyzer with Agglutinative Affix Morphemes and Fusional Concatenation Rules”, COLING, Mumbai, India, December, 2012b.
  45. Zaraket F., Makhlouta J., “Arabic Temporal Entity Extraction using Morphological Analysis”, IJCLA, vol. 3, p. 121-136, 2012c. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/jaber-2017-morphology-based-entity-relational-entity/