Elamur Research⌕
☰
home / papers / 2020 / Arabic Curriculum Analysis

Arabic Curriculum Analysis

Paper · 2020 · Hamdy Mubarak, Shimaa Amer, Ahmed Abdelali et al.

Summary

Arabic has few NLP tools for checking whether school textbooks meet their learning goals, for example in the vocabulary they teach. The authors built a public web system (curriculum.qcri.org) to explore this. It covers primary-school Arabic-subject textbooks for grades 1–6 from six Gulf countries. The texts were processed with the Farasa toolkit for segmentation, diacritization and lemmatization, and language experts sorted lemmas into 50 topical categories.

Users can:

  • see how often a word and its inflected forms appear by grade and country;
  • browse lemmas by category;
  • view per-grade or cumulative lemma statistics;
  • list vocabulary found only in one country's curriculum;
  • "grade" any input text by flagging words whose lemmas never occur in the chosen grade or earlier ones.

Text-to-speech, machine translation and morphological analysis are added to help teachers and learners. The statistics show vocabulary growing from about 1.5k lemmas in grade 1 to about 8k by grade 6. The Qatari and Omani curricula are lexically poorer than the others. The tool gives curriculum designers a data-driven view of coverage and difficulty.

Key points

  • The corpus comes from 2014/2015 textbooks from BH, KW, OM, QA, SA and AE. In total it has about 166K sentences, 895K tokens and 10.5K unique lemmas. AE grades 3 and 5 are incomplete.
  • Processing uses Farasa for morphological segmentation, diacritization and lemmatization. Experts manually assigned lemmas to 50 categories, such as Function Words, Animal, Food and Politics.
  • Five features are offered (Text Grading, Term Usage, Category, Differences and Statistics), and the system runs on Django following an MVC design.
  • Cumulative vocabulary is about 1.5k lemmas after grade 1 and about 8k after grade 6. QA and OM have lower vocabulary richness than the other countries.
  • Text Grading marks a word as difficult when its lemma has not appeared in the selected or any earlier grade. Fewer words are flagged as the chosen grade rises.

Interesting ideas

  • I1. A cumulative, grade-indexed lemma inventory works as a simple and fully explainable difficulty oracle: a word is "hard" exactly when it has not yet been taught.
  • I2. Counting at the lemma level while still showing the spread of surface forms per lemma separates "knowing a word" from "knowing its inflections."
  • I3. Expert-assigned topical categories over lemmas let users measure how well a corpus covers a theme, such as animals or food.
  • I4. A "unique to this collection" view contrasts one corpus's vocabulary against several sibling corpora to expose gaps and peculiarities.
  • I5. The grade-level vocabulary check doubles as a constraint for producing simpler text, which the authors list as future work on simplification.

For Mova

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

  • I1 (answer): School-course test readers could check whether a question's vocabulary lies within the grade it targets. They could also explain a failure as an unseen-lemma gap rather than a reasoning error.
  • I2 (tokens): Per-lemma distributions of inflected forms in a domain corpus could show which morphological forms the tagger and lemmatizer most need to handle well.
  • I3 (world): Mapping lemmas to ontology classes from WordNet or Wikidata would give coverage reports per topic for the compiled knowledge layer, showing domains where entities or verb classes are thin.
  • I4 (learn): Comparing the vocabulary of silver training data against evaluation sets could reveal lemmas the models never saw, which would guide targeted data collection.
  • I5 (back): Generation from the meaning graph could limit word choice to a chosen grade's lemma inventory, giving controllable, age-appropriate retellings of fables and tales.

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

🎓Hamad bin Khalifa UniversityDoha · Qatar144 works · 14,175 citations🎓Qatar UniversityDoha · Qatar53 works · 1,027 citations

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksSocial Sciences · 1,954 worksEducation · 212 works

🧭 Topics 1Research subjects and languages the work is about.Semitic languages · 3,066 works

Cite

Cite as Hamdy Mubarak, Shimaa Amer, Ahmed Abdelali and Kareem Darwish. 2020. Arabic Curriculum Analysis. Proceedings of the 28th International Conference on Computational Linguistics: System Demonstrations.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2020.coling-demos.15

@inproceedings{mubarak2020arabic,
  title = {{Arabic Curriculum Analysis}},
  author = {Hamdy Mubarak, Shimaa Amer, Ahmed Abdelali and Kareem Darwish},
  booktitle = {Proceedings of the 28th International Conference on Computational Linguistics: System Demonstrations},
  year = {2020},
  doi = {10.18653/v1/2020.coling-demos.15},
  url = {https://doi.org/10.18653/v1/2020.coling-demos.15}
}

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. Ahmed Abdelali, Kareem Darwish, Nadir Durrani, and Hamdy Mubarak. 2016. Farasa: A fast and furious segmenter for arabic. In Proceedings of the 2016 conference of the North American chapter of the association for computational linguistics: Demonstrations, pages 11–16.
  2. Hend S Al-Khalifa and Amani A Al-Ajlan. 2010. Automatic readability measurements of the arabic text: An exploratory study. Arabian Journal for Science and Engineering, 35(2 C):103–124. find
  3. Muhamed Al Khalil, Hind Saddiki, Nizar Habash, and Latifa Alfalasi. 2018. A leveled reading corpus of modern standard arabic. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan, May. European Language Resources Association (ELRA).
  4. Rebekah George Benjamin. 2012. Reconstructing readability: Recent developments and recommendations in the analysis of text difficulty. Educational Psychology Review, 24(1):63–88. ↗
  5. Sabine Braun, Kurt Kohn, and J Mukherjee. 2006. Corpus Technology and Language Pedagogy: New Resources, New Tools, New Methods. English Corpus Linguistics Vol 3. Lang. find
  6. Kevyn Collins-Thompson. 2014. Computational assessment of text readability: A survey of current and future research. ITL - International Journal of Applied Linguistics, 165(2):97–135. find
  7. Kareem Darwish, Hamdy Mubarak, and Ahmed Abdelali. 2017. Arabic diacritization: Stats, rules, and hacks. In Proceedings of the Third Arabic Natural Language Processing Workshop, pages 9–17, Valencia, Spain, April. Association for Computational Linguistics.
  8. Marcos Garcı́a Salido, Marcos Garcia, Milka Villayandre-Llamazares, and Margarita Alonso-Ramos. 2018. A lexical tool for academic writing in spanish based on expert and novice corpora. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Osaka, Japan, May. European Language Resources Association (ELRA). find
  9. Hamdy Mubarak. 2018. Build fast and accurate lemmatization for Arabic. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan, May. European Language Resources Association (ELRA).
  10. Peter F Oliva, 2005. Developing the curriculum (6th ed.). Pearson, Boston. find
  11. Sandy Soto. 2015. An analysis of curriculum development. Theory and Practice in Language Studies, 5:1129, 06. ↗
  12. R. W. Tyler. 1950. Basic Principles of Curriculum and Instruction. University of Chicago Press. find
  13. Wajdi Zaghouani, Behrang Mohit, Nizar Habash, Ossama Obeid, Nadi Tomeh, Alla Rozovskaya, Noura Farra, Sarah Alkuhlani, and Kemal Oflazer. 2014. Large scale arabic error annotation: Guidelines and framework. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14), Reykjavik, Iceland, may. European Language Resources Association (ELRA). ↗
  14. Nasser Zalmout, Hind Saddiki, and Nizar Habash. 2016. Analysis of foreign language teaching methods: An automatic readability approach. In Proceedings of the 3rd Workshop on Natural Language Processing Techniques for Educational Applications (NLPTEA2016), pages 122–130, Osaka, Japan, December. find
  15. The COLING 2016 Organizing Committee. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/mubarak-2020-arabic-curriculum-analysis/