Elamur Research⌕
☰
home / papers / 2012 / Developing ARET: An NLP-based Educational Tool Set for Arabic Reading Enhancement

Developing ARET: An NLP-based Educational Tool Set for Arabic Reading Enhancement

Paper · 2012 · Mohammed Maamouri, Wajdi Zaghouani, Violetta Cavalli-Sforza et al.

Summary

Learners of Modern Standard Arabic struggle to read because written Arabic usually omits short vowels and other diacritics, and its morphology is rich. A single undiacritized form can therefore have many readings. The authors at LDC built ARET, a web-based classroom tool set with two parts: ARFT, a facilitation tool for browsing texts, and ARAT, an assessment tool. Its content is the full set of reading passages from the popular Al-Kitaab textbook series, about 60,000 words. Each word was analyzed by SAMA, which lists every candidate analysis. A native-speaker annotator then picked the analysis that fit the context, and expert linguists reviewed the result. The output is a fully diacritized, segmented, POS-tagged and glossed corpus stored in a relational database. The database supports concordances, lemma dictionary lookups, and spoken renderings through commercial TTS with male and female voices. The system also generates test questions with distractors. The paper reports no classroom impact evaluation yet. Its value lies in the reusable annotated resource and a pipeline that was later ported to Nahuatl.

Key points

  • The corpus database holds 3,692 sentences or headings, 53,411 word tokens, 17,209 undiacritized and 20,725 diacritized forms, 22,304 distinct analyses, 6,829 lemmas, and 3,436 glossary entries.
  • Annotation followed the Arabic Treebank recipe: SAMA lists all solutions, a human selects one in context, and gaps in POS tags or glosses are edited by hand.
  • Separate relational tables cover passages, sentences, tokens, word types with context-dependent analyses, and lemmas plus clitics. Together they let the system rebuild any passage and link each token to its lemma's dictionary entry.
  • The authors added 1,022 tokens of news text as material students have not seen. These texts are usable in ARAT quizzes but cannot be browsed in ARFT.
  • At the time of writing, 118 University of Pennsylvania students were registered. The authors originally planned to adapt REAP but had to build their own system instead.

Interesting ideas

  • I1. Annotators chose among an analyzer's full candidate list instead of labeling from scratch, and edited only where the analyzer's coverage fell short; this makes annotation faster and more consistent.
  • I2. Tokens point to a shared inventory of word types with context-dependent analyses, and those types point to lemma entries; this normalized layout gives concordances and dictionary lookups almost for free.
  • I3. The authors kept a separately annotated set of texts that is available for testing but hidden from browsing, so quizzes measure reading of truly new material.
  • I4. Quiz questions and distractors come from the same annotated corpus that students read, so assessment is tied directly to the study material.
  • I5. One undiacritized lemma maps to nine readings, which shows that word recognition really means resolving among enumerated alternatives in context.

For Mova

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

  • I1 (learn): an LLM teacher could be shown the candidate analyses proposed by the witness parsers and asked to pick or minimally fix one, rather than produce a free-form parse.
  • I2 (tokens): crate en could store each disambiguated analysis once and have tokens reference it, so every lemma and morphology decision can be traced and searched across texts.
  • I3 (answer): the school-course and fable QA readers could be scored on passages kept apart from any material used for tuning, as a guard against contamination.
  • I4 (answer): Mova could build self-test questions, including plausible wrong options, from its own world-of-the-text graph to check whether its answers are consistent.
  • I5 (tokens): the tagger could keep an explicit list of readings for English homographs and record which piece of context ruled each one out, so its choices stay explainable.

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 PennsylvaniaPhiladelphia · United States221 works · 63,234 citations🎓Al Akhawayn UniversityIfrane · Morocco3 works · 247 citations

 Fields, topics and concepts

📚 Shelves 8The 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 worksSocial Sciences · 1,772 worksPsychology · 1,552 worksDevelopmental and Educational Psychology · 499 worksEducation · 203 works

🧭 Topics 1Research subjects and languages the work is about.Semitic languages · 2,908 works

Cite

Cite as Mohammed Maamouri, Wajdi Zaghouani, Violetta Cavalli-Sforza, Dave Graff and Mike Ciul. 2012. Developing ARET: An NLP-based Educational Tool Set for Arabic Reading Enhancement. Proceedings of the Seventh Workshop on Building Educational Applications Using NLP.

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

@inproceedings{maamouri2012developing,
  title = {{Developing ARET: An NLP-based Educational Tool Set for Arabic Reading Enhancement}},
  author = {Mohammed Maamouri, Wajdi Zaghouani, Violetta Cavalli-Sforza, Dave Graff and Mike Ciul},
  booktitle = {Proceedings of the Seventh Workshop on Building Educational Applications Using NLP},
  year = {2012},
  url = {https://aclanthology.org/W12-2015/}
}

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. Mahmoud Al-Batal, Kristen Brustad & Abbas Al-Tonsi. 2006. Al-Kitaab fii tacallum al-cArabiyya, Volume II (with DVDs, Second Edition).Washington, D.C.: Georgetown University Press, 2006. find
  2. Mahmoud Al-Batal, Kristen Brustad & Abbas Al-Tonsi. 2004. Al-Kitaab fii tacallum al-cArabiyya, A Textbook for Beginning Arabic, Volume I (with DVDs, Second Edition). Washington, D.C.: Georgetown University Press, 2004. find
  3. Mahmoud Al-Batal, Kristen Brustad & Abbas Al-Tonsi. 2001. Al-Kitaab fii tacallum al-cArabiyya,Volume III.Washington, D.C.: Georgetown University Press, 2001. find
  4. Jon Brown and Maxine Eskenazi. 2004. Retrieval of authentic documents for reader-specific lexical practice. In Proceedings of InSTIL/ICALL Symposium 2004. Venice, Italy. find
  5. Tsuo-Lin Chiu, Hsien-Chin Liou and Yuli Yeha. 2007. A study of web-based oral activities enhanced byau-tomatic speech recognition for EFL college learning. Computer Assisted Language Learning, 20 (3), 209– 233. find
  6. Michael Heilman, Kevyn Collins-Thompson, Jamie Callan, and Maxine Eskenazi. 2006. Classroom success of an Intelligent Tutoring System for lexical practice and reading comprehension. In Proceedings of the Ninth International Conference on Spoken LanguageProcessing. Pittsburgh, PA. find
  7. Mohamed Maamouri, David Graff, Basma Bouziri, Sondos Krouna, Ann Bies and Seth Kulick. 2010. Standard Arabic Morphological Analyzer (SAMA) Version 3.1. Linguistic Data Consortium, Catalog No.: LDC2010L01. find
  8. Mohamed Maamouri and Ann Bies. 2004. Developing an Arabic Treebank: Methods, Guidelines, Procedures, and Tools. In Proceedings of the Workshop Computational Approaches to Arabic Script-based Languages. Pages 2-9./20th International Conference on Computational Linguistics/. COLING Geneva, Switzerland.
  9. Detmar Meurers, Ramon Ziai, Luiz Amaral, Adriane Boyd, Aleksandar Dimitrov, Vanessa Metcalf, Niels Ott. 2010. Enhancing Authentic Web Pages for Language Learners. In Proceedings of the 5th Workshop on Innovative Use of NLP for Building Educational Applications, NAACL-HLT 2010, Los Angeles. find
  10. Roland Meynet. 1971. L'écriture arabe en question: les projets de l'Académie de Langue Arabe du Caire de 1938 à 1968.Beirut: Dar el-Machreq, 1971. 142 pp Eleni Miltsakaki and Audrey Troutt. 2007. Read-X: Automatic Evaluation of Reading Difficulty of Web Text. In Proceedings of E-Learn 2007, sponsored by the Association for theAdvancement of Computing in Education. Quebec, Canada. find
  11. John Nerbonne and Petra Smit. 1996. GLOSSER-RuG: in Support of Reading. In Proceedings of the 16th International Conference on Computational Linguistics (COLING 1996). find
  12. Kathleen M. Sheehan, Irene Kostin, Yoko Futagi. 2007. SourceFinder: A Construct-Driven Approach for Locating Appropriately Targeted Reading Comprehension Source Texts. In Proceedings of the SLaTE Workshop on Speech and Language Technology in Education.Carnegie Mellon University and International Speech Communication Association (ISCA). find
  13. Li Wang. 2005. The advantages of using technology in second language education. T.H.E. Journal, 32 (10), 1-6. find
  14. Paige D. Ware. 2004. Confidence and competition online: ESL student perspectives on web based discussions in the classroom. Computers and Composition, 21, 451–468. ↗
  15. Yi Yuan. 2003. The use of chat rooms in an ESL setting. Computers and Composition, 20, 194–206. ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/maamouri-2012-developing-aret-nlp-based-educational/