Elamur Research⌕
☰
home / papers / 2026 / Masrad: Arabic Terminology Management Corpora with Semi-Automatic Construction

Masrad: Arabic Terminology Management Corpora with Semi-Automatic Construction

Summary

Arabic translations of specialized and academic works often render the same foreign concept with different Arabic terms. Translators vary, and no reference collection of parallel terms exists. The authors address this by building Masrad, a termbase of foreign–Arabic term pairs drawn from Arabic books published by ACRPS. They exploit a common habit of Arabic scholarly writing: a foreign term is placed in parentheses right after its Arabic rendering. The difficulty is deciding how many of the preceding Arabic words belong to the term. Their tool, Masrad-Ex, generates every candidate span of increasing length before the parenthesis. It then scores each candidate with lexicographic, phonetic, morphological and semantic similarity features. Heuristic, machine-learning and ML-with-post-processing variants pick the best span. The strongest variant reaches 90.5% precision and 92.4% recall. The system runs locally without any LLM, so manuscripts are not exposed to outside services. Comparison with an existing in-house glossary shows that most extracted terms are new. This makes the resource useful for editors who need consistent terminology.

Key points

  • Full processed collection: 495 books, 58,570 unique foreign terms, 84,242 term occurrences and 334,564 candidate Arabic spans, about 3.97 candidates per occurrence.
  • Annotated Masrad subset: 15 books, 4,347 unique foreign terms, 4,841 occurrences and 19,405 labeled candidates. Labels were drafted with heuristics and then corrected by experts.
  • Comparison with the in-house glossary: the glossary holds 46,787 English/French/Arabic concepts. Only 7,368 of them (15.74%) matched extracted terms, and at least 74.84% of the extracted terms appear to be novel.
  • Availability: Masrad is shared with researchers on request. The training/testing feature vectors are published online.
  • Scope and future work: the tool is meant to support expert judgment, not replace it. Planned work includes fine-tuning a sentence transformer for the semantic feature and pruning or adding features.

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: Linguistics & languages

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/4azqkq8r4eez

@inproceedings{nasser2026masrad,
  title = {{Masrad: Arabic Terminology Management Corpora with Semi-Automatic Construction}},
  author = {Nasser, Mahdi and Sayah, Laura and Zaraket, Fadi},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/4azqkq8r4eez},
  url = {https://doi.org/10.63317/4azqkq8r4eez}
}

References

17 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ahmet Aker, Monica Paramita, and Rob Gaizauskas. 2013. Extracting bilingual terminologies from comparable corpora. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 402–411, Sofia, Bulgaria. Association for Computational Linguistics. Nidhal Baccouri. deep-translator. find
  2. Ido Dagan and Ken Church. 1994. Termight: Identifying and translating technical terminology. In Fourth Conference on Applied Natural Language Processing, pages 34–40, Stuttgart, Germany. Association for Computational Linguistics. find
  3. Fangxiaoyu Feng, Yinfei Yang, Daniel Cer, Naveen Arivazhagan, and Wei Wang. 2022. Language-agnostic bert sentence embedding. find our card
  4. Eibe Frank, Mark A. Hall, and Ian H. Witten. 2016. The WEKA Workbench. Online Appendix for "Data Mining: Practical Machine Learning Tools and Techniques", fourth edition. Morgan Kaufmann. URL: http://www.cs.waikato.ac.nz/ml/weka/Witten_et_al_2016_appendix.pdf. GokulNC and Prem27. google-transliteration-api. ↗
  5. Matthew Honnibal, Ines Montani, Sofie Van Lan-deghem, and Adriane Boyd. 2020. spacy: Industrial-strength natural language processing in python. find
  6. Mustafa Jarrar, Mohammed Khalilia, and Sana Ghanem. 2022. Wojood: Nested arabic named entity corpus and recognition using bert. find
  7. Vladimir I. Levenshtein. 1966. Binary codes capable of correcting deletions, insertions, and reversals. Soviet Physics Doklady, 10:707–710. find
  8. Nils Reimers and Iryna Gurevych. 2019. Sentence-bert: Sentence embeddings using siamese bert-networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics. find
  9. Robert C. Russell and Margaret King Odell. 1918. System of soundexing. https://patents.google.com/patent/US1261167A/en. ↗
  10. Gerard Salton, Anita Wong, and Chung-Shu Yang. 1975. A vector space model for automatic indexing. Communications of the ACM, 18(11):613– 620. ↗
  11. Awara Sharif. http://awara85.blogspot.com/p/blog-page.html. Accessed: 2024-05-16. 8. ↗
  12. Language Resource References Maged Saeed Al-Shaibani. 2021. MagedSaeed/farasapy: Python wrapper for the Farasa toolkit. PID https://github.com/MagedSaeed/farasapy. Accessed: 2025-07-07. ↗
  13. Global WordNet Association. Arabic WordNet - Arabic Resources. PID https://globalwordnet.github.io/resources/wordnets-in-the-world. Accessed: 2025-10-25. ↗
  14. Mustafa Jarrar. 2022. The Arabic Ontology - An Arabic Wordnet with Ontologically Clean Content. PID https://ontology.birzeit.edu/. ↗
  15. Hossam Mahdy. Glossary of Arabic Terms for the Conservation of Cultural Heritage in Arabic Alphabetical Order. PID https://www.iccrom.org/sites/default/files/2017-12/iccrom g lossary e n − ar.pd f . Accessed: 2025-10-25. ↗
  16. Nagoudi, El Moatez Billah and Abdul-Mageed, Muhammad and Bouamor, Houda and Habash, Nizar. 2022. TURJUMAN: A Public Toolkit for Neural Arabic Machine Translation. PID https://github.com/UBC-NLP/turjuman. ↗
  17. Obeid, Ossama and Zalmout, Nasser and Khalifa, Salam and Taji, Dima and Oudah, Mai and Alhafni, Bashar and Inoue, Go and Eryani, Fadhl and Erdmann, Alexander and Habash, Nizar. 2020. CAMeL Tools: An Open Source Python Toolkit for Arabic Natural Language Processing. European Language Resources Association. PID https://www.aclweb.org/anthology/2020.lrec-1.868. ↗ our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/nasser-2026-masrad-arabic-terminology-management-corpora/