Elamur Research⌕
☰
home / papers / 2026 / ArabiGEE: A Hierarchical Taxonomy for Arabic Grammatical Error Explanation

ArabiGEE: A Hierarchical Taxonomy for Arabic Grammatical Error Explanation

Summary

Explanations of Arabic grammar errors have so far been produced as free text. That makes them hard to score automatically, and older Arabic error taxonomies such as ALC are too flat to support them. The authors present ArabiGEE, a hand-built hierarchy for Arabic grammatical error explanation (GEE) that is organized into four dimensions, namely the orthographic, morphological, syntactic and lexical levels. Within every one of these, the hierarchy branches from error types down to correction types and finally to explanation types. Every explanation type is tied to one fixed explanation text written for learners. A single correction can receive a group of explanations from several dimensions. For example, a hamza spelling fix may rest on case morphology, and the case in turn comes from the syntactic role of the word. The team used the scheme to annotate 1,000 word pairs drawn from existing Arabic GEC corpora. They then evaluated LLMs by asking them to choose labels from the taxonomy. Frontier models such as GPT-5.5 and Gemini perform reasonably well but struggle with the finest distinctions. This turns GEE into a task that can be measured.

Key points

  • The taxonomy has 27 error types, 140 correction types and 324 explanation types. By dimension (error / correction / explanation): Lexical 2/7/10, Orthographic 5/54/86, Morphological 13/60/104, Syntactic 7/19/124.
  • ALC labels a word with two separate hamza errors under one Hamza tag. ArabiGEE separates the two instances and can explain one fix across orthography, morphology and syntax at once.
  • The release contains the taxonomy plus 1,000 manually annotated word pairs taken from existing Arabic GEC data, together with the code.
  • GPT and Gemini were the strongest models. Lexical explanations reach about 80 F1, while morphology never goes above 67 F1. Dropping the corrected-sentence context hurts syntactic explanations the most.
  • The error analysis of GPT-5.5 shows that most failures are spurious or missing explanations rather than confused labels. Morphology had only 4 wrong labels but 67 extra predictions. Most label confusions differed from gold only at the finest subtype level, and 8 gold annotations turned out to be wrong.

Interesting ideas

  • I1. Each fine explanation type maps to one fixed explanation text, so producing an explanation becomes choosing a label that can be scored exactly instead of compared as free text. — For Mova (answer): Students could justify each answer by citing a closed inventory of reasons, which lets us grade both the answer and the stated reason.
  • I2. Several linked explanations across dimensions can jointly justify one correction, as in a spelling form that depends on case, which depends on the syntactic role. — For Mova (tokens): A morphological feature such as case or number could store a pointer to the UD relation that licenses it, which makes tagger decisions explainable through the tree.
  • I3. Predictions are sorted into matched-correct, matched-wrong, extra and missing after alignment, which separates "found the wrong spots" from "named the right spot wrongly." — For Mova (learn): The same four buckets would help when comparing witness parsers or frame extractors against gold, showing whether repair rules should target detection or labeling.
  • I4. Removing part of the input, here the corrected sentence, reveals which dimension actually depends on that context. — For Mova (learn): This works as a negative control: strip context from a test item and check that context-dependent stages degrade while others stay stable.
  • I5. Scoring at every level of the hierarchy shows that many errors are near-misses that land in the right region of the taxonomy. — For Mova (frames): Scoring VerbNet role or class predictions level by level would credit predictions that are close but too coarse, and point to where finer distinctions fail.

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 10.48550/arxiv.2606.10765 · arXiv 2606.10765

@misc{elhady2026arabigee,
  title = {{ArabiGEE: A Hierarchical Taxonomy for Arabic Grammatical Error Explanation}},
  author = {Elhady, Khaled and Kallas, Omar and Habash, Nizar and Alhafni, Bashar},
  year = {2026},
  doi = {10.48550/arxiv.2606.10765},
  eprint = {2606.10765},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2606.10765}
}

References

52 references, 13 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Abdullah Alfaifi and Eric Atwell. 2012. Arabic learner corpora (ALC): A taxonomy of coding errors. In The 8th International Computing Conference in Arabic. find
  2. Abdullah Alfaifi and Eric Atwell. 2014. An evaluation of the Arabic error tagset v2. In Proceedings of the AACL 2014-The American Association for Corpus Linguistics conference. The American Association for Corpus Linguistics. find
  3. Abdullah Alfaifi, Eric Atwell, and Ghazi Abuhakema. 2013. Error annotation of the Arabic learner corpus. In Language Processing and Knowledge in the Web, pages 14–22. Springer. find
  4. Bashar Alhafni and Nizar Habash. 2025. Enhancing text editing for grammatical error correction: Arabic as a case study. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 17892–17914, Vienna, Austria. Association for Computational Linguistics. find
  5. Bashar Alhafni, Go Inoue, Christian Khairallah, and Nizar Habash. 2023. Advancements in Arabic grammatical error detection and correction: An empirical investigation. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 6430–6448, Singapore. Association for Computational Linguistics. find our card
  6. Anthropic. 2025. System card: Claude opus 4 & claude sonnet 4. Technical report, Anthropic. find
  7. Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, Binyuan Hui, Luo Ji, Mei Li, Junyang Lin, Runji Lin, Dayiheng Liu, Gao Liu, Chengqiang Lu, Keming Lu, and 29 others. 2023. Qwen technical report. Preprint, arXiv:2309.16609. ↗ our card
  8. M Saiful Bari, Yazeed Alnumay, Norah A. Alzahrani, Nouf M. Alotaibi, Hisham Abdullah Alyahya, Sultan AlRashed, Faisal Abdulrahman Mirza, Shaykhah Z. Alsubaie, Hassan A. Alahmed, Ghadah Alabduljabbar, Raghad Alkhathran, Yousef Almushayqih, Raneem Alnajim, Salman Alsubaihi, Maryam Al Mansour, Saad Amin Hassan, Dr. Majed Alrubaian, Ali Alammari, Zaki Alawami, and 7 others. 2025. AL-Lam: Large language models for arabic and english. In The Thirteenth International Conference on Learning Representations. find
  9. Riadh Belkebir and Nizar Habash. 2021. Automatic error type annotation for Arabic. In Proceedings of the 25th Conference on Computational Natural Language Learning, pages 596–606, Online. Association for Computational Linguistics. find our card
  10. Adriane Boyd, Jirka Hana, Lionel Nicolas, Detmar Meurers, Katrin Wisniewski, Andrea Abel, Karin Schöne, Barbora Štindlová, and Chiara Vettori. 2014. The MERLIN corpus: Learner language and the CEFR. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14), pages 1281–1288, Reykjavik, Iceland. European Language Resources Association (ELRA). find
  11. Dan Brown, Qiandi Liu, and Reza Norouzian. 2023. Effectiveness of written corrective feedback in developing l2 accuracy: A bayesian meta-analysis. Language Teaching Research, 30(3):1357–1389. find
  12. Christopher Bryant, Mariano Felice, and Ted Briscoe. 2017. Automatic annotation and evaluation of error types for grammatical error correction. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 793–805, Vancouver, Canada. Association for Computational Linguistics. find our card
  13. Christopher Bryant, Zheng Yuan, Muhammad Reza Qorib, Hannan Cao, Hwee Tou Ng, and Ted Briscoe. 2023. Grammatical error correction: A survey of the state of the art. Computational Linguistics, 49(3):643–701. find
  14. Leshem Choshen, Dmitry Nikolaev, Yevgeni Berzak, and Omri Abend. 2020. Classifying syntactic errors in learner language. In Proceedings of the 24th Conference on Computational Natural Language Learning, pages 97–107, Online. Association for Computational Linguistics. find our card
  15. Steven Coyne, Diana Galvan-Sosa, Ryan Spring, Camélia Guerraoui, Michael Zock, Keisuke Sakaguchi, and Kentaro Inui. 2025. Annotating errors in english learners’ written language production: Advancing automated written feedback systems. In Artificial Intelligence in Education, pages 292–306, Cham. Springer Nature Switzerland. find
  16. Daniel Dahlmeier, Hwee Tou Ng, and Siew Mei Wu. 2013. Building a large annotated corpus of learner English: The NUS corpus of learner English. In Proceedings of the Eighth Workshop on Innovative Use of NLP for Building Educational Applications, pages 22–31, Atlanta, Georgia. Association for Computational Linguistics. find
  17. DeepSeek-AI, Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, and 181 others. 2025. Deepseek-v3 technical report. Preprint, arXiv:2412.19437. ↗
  18. Mohammed Elkholy, Khalid N. Elmadani, Nizar Habash, and Bashar Alhafni. 2026. Arabic sentence segmentation across genres and punctuation conditions. Preprint, arXiv:2606.08025. ↗ our card
  19. Charles F Ferguson. 1959. Diglossia. Word, 15(2):325– 340. find
  20. Nizar Habash, Mona Diab, and Owen Rambow. 2012. Conventional orthography for dialectal Arabic. In Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC’12), pages 711–718, Istanbul, Turkey. European Language Resources Association (ELRA). find our card
  21. Nizar Habash and David Palfreyman. 2022. ZAEBUC: An annotated Arabic-English bilingual writer corpus. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 79–88, Marseille, France. European Language Resources Association. find
  22. Nizar Habash, Owen Rambow, Mona Diab, and Reem Kanjawi-Faraj. 2008. Guidelines for Annotation of Arabic Dialectness. In Proceedings of the Workshop on HLT & NLP within the Arabic World, Marrakech, Morocco. find
  23. Nizar Habash, Abdelhadi Soudi, and Tim Buckwalter. 2007. On Arabic Transliteration. In A. van den Bosch and A. Soudi, editors, Arabic Computational Morphology: Knowledge-based and Empirical Methods, pages 15–22. Springer, Netherlands. find
  24. Nizar Y Habash. 2010. Introduction to Arabic natural language processing, volume 3. Morgan & Claypool Publishers. find
  25. Albert Q. Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, Lélio Renard Lavaud, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, and William El Sayed. 2023. Mistral 7b. Preprint, arXiv:2310.06825. ↗
  26. Masahiro Kaneko and Naoaki Okazaki. 2024. Controlled generation with prompt insertion for natural language explanations in grammatical error correction. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 3955–3961, Torino, Italia. ELRA and ICCL. find
  27. EunYoung Kang and Zhaohong Han. 2015. The efficacy of written corrective feedback in improving L2 written accuracy: A meta-analysis. The Modern Language Journal, 99(1):1–18. find
  28. S. Magalí López Cortez, Mark Josef Norris, and Steve Duman. 2024. GMEG-EXP: A dataset of human-and LLM-generated explanations of grammatical and fluency edits. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 7785–7800, Torino, Italia. ELRA and ICCL. find
  29. Xiwen Lu, Qingyu Yang, Jian Wei, and Neil T. Heffernan. 2025. From belief to evidence: simpler immediate feedback improves language learning and confidence in semi-open-ended questions. Frontiers in Education, 10. find
  30. Samar Mohamed Magdy, Fakhraddin Alwajih, Sang Yun Kwon, Reem Abdel-Salam, and Muhammad Abdul-Mageed. 2024. Gazelle: An instruction dataset for Arabic writing assistance. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 16027–16054, Miami, Florida, USA. Association for Computational Linguistics. find
  31. Subhankar Maity, Aniket Deroy, and Sudeshna Sarkar. 2024. How ready are generative pre-trained large language models for explaining bengali grammatical errors? Preprint, arXiv:2406.00039. ↗
  32. Behrang Mohit, Alla Rozovskaya, Nizar Habash, Wajdi Zaghouani, and Ossama Obeid. 2014. The first QALB shared task on automatic text correction for Arabic. In Proceedings of the EMNLP 2014 Workshop on Arabic Natural Language Processing (ANLP), pages 39–47, Doha, Qatar. Association for Computational Linguistics. find
  33. Hamdy Mubarak, Majd Hawasly, and Abubakr Mohamed. 2026. Nahw: A comprehensive benchmark of Arabic grammar understanding, error detection, correction, and explanation. In Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 6310–6328, Rabat, Morocco. Association for Computational Linguistics. find our card
  34. Ryo Nagata. 2019. Toward a task of feedback comment generation for writing learning. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 3206–3215, Hong Kong, China. Association for Computational Linguistics. find
  35. Ryo Nagata, Masato Hagiwara, Kazuaki Hanawa, Masato Mita, Artem Chernodub, and Olena Nahorna. 2021. Shared task on feedback comment generation for language learners. In Proceedings of the 14th International Conference on Natural Language Generation, pages 320–324, Aberdeen, Scotland, UK. Association for Computational Linguistics. find
  36. Ryo Nagata, Kentaro Inui, and Shin’ichiro Ishikawa. 2020. Creating corpora for research in feedback comment generation. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 340–345, Marseille, France. European Language Resources Association. find
  37. Jakub Náplava, Milan Straka, Jana Straková, and Alexandr Rosen. 2022. Czech grammar error correction with a large and diverse corpus. Transactions of the Association for Computational Linguistics, 10:452–467. find
  38. Hwee Tou Ng, Siew Mei Wu, Ted Briscoe, Christian Hadiwinoto, Raymond Hendy Susanto, and Christopher Bryant. 2014. The CoNLL-2014 shared task on grammatical error correction. In Proceedings of the Eighteenth Conference on Computational Natural Language Learning: Shared Task, pages 1–14, Baltimore, Maryland. Association for Computational Linguistics. find
  39. Hwee Tou Ng, Siew Mei Wu, Yuanbin Wu, Christian Hadiwinoto, and Joel Tetreault. 2013. The CoNLL-2013 shared task on grammatical error correction. In Proceedings of the Seventeenth Conference on Computational Natural Language Learning: Shared Task, pages 1–12, Sofia, Bulgaria. Association for Computational Linguistics. find
  40. Alla Rozovskaya, Houda Bouamor, Nizar Habash, Wajdi Zaghouani, Ossama Obeid, and Behrang Mohit. 2015. The second QALB shared task on automatic text correction for Arabic. In Proceedings of the Second Workshop on Arabic Natural Language Processing, pages 26–35, Beijing, China. Association for Computational Linguistics. find
  41. Alla Rozovskaya and Dan Roth. 2019. Grammar error correction in morphologically rich languages: The case of Russian. Transactions of the Association for Computational Linguistics, 7:1–17. find
  42. Neha Sengupta, Sunil Kumar Sahu, Bokang Jia, Satheesh Katipomu, Haonan Li, Fajri Koto, William Marshall, Gurpreet Gosal, Cynthia Liu, Zhiming Chen, Osama Mohammed Afzal, Samta Kamboj, Onkar Pandit, Rahul Pal, Lalit Pradhan, Zain Muhammad Mujahid, Massa Baali, Xudong Han, Sondos Mahmoud Bsharat, and 13 others. 2023. Jais and jais-chat: Arabic-centric foundation and instruction-tuned open generative large language models. Preprint, arXiv:2308.16149. ↗ our card
  43. Aaditya Singh, Adam Fry, Adam Perelman, Adam Tart, Adi Ganesh, Ahmed El-Kishky, Aidan McLaughlin, Aiden Low, AJ Ostrow, Akhila Ananthram, Akshay Nathan, Alan Luo, Alec Helyar, Aleksander Madry, Aleksandr Efremov, Aleksandra Spyra, Alex Baker-Whitcomb, Alex Beutel, Alex Karpenko, and 467 others. 2026. Openai gpt-5 system card. Preprint, arXiv:2601.03267. ↗ our card
  44. Yixiao Song, Kalpesh Krishna, Rajesh Bhatt, Kevin Gimpel, and Mohit Iyyer. 2024. GEE! grammar error explanation with large language models. In Findings of the Association for Computational Linguistics: NAACL 2024, pages 754–781, Mexico City, Mexico. Association for Computational Linguistics. find
  45. Fanar Team, Ummar Abbas, Mohammad Shahmeer Ahmad, Firoj Alam, Enes Altinisik, Ehsannedin Asgari, Yazan Boshmaf, Sabri Boughorbel, Sanjay Chawla, Shammur Chowdhury, Fahim Dalvi, Kareem Darwish, Nadir Durrani, Mohamed Elfeky, Ahmed Elmagarmid, Mohamed Eltabakh, Masoomali Fatehkia, Anastasios Fragkopoulos, Maram Hasanain, and 23 others. 2025a. Fanar: An arabic-centric multimodal generative ai platform. find our card
  46. Gemini Team, Rohan Anil, Sebastian Borgeaud, Jean-Baptiste Alayrac, Jiahui Yu, Radu Soricut, Johan Schalkwyk, Andrew M. Dai, Anja Hauth, Katie Millican, David Silver, Melvin Johnson, Ioannis Antonoglou, Julian Schrittwieser, Amelia Glaese, Jilin Chen, Emily Pitler, Timothy Lillicrap, Angeliki Lazaridou, and 1 others. 2025b. Gemini: A family of highly capable multimodal models. Preprint, arXiv:2312.11805. ↗ our card
  47. Martin Vainikko, Taavi Kamarik, Karina Kert, Krista Liin, Silvia Maine, Kais Allkivi, Annekatrin Kaivapalu, and Mark Fishel. 2025. Paragraph-level error correction and explanation generation: Case study for Estonian. In Proceedings of the 20th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2025), pages 953–967, Vienna, Austria. Association for Computational Linguistics. find
  48. Reina Wakabayashi. 2008. The effect of peer feedback on EFL writing: Focusing on Japanese university students. OnCUE Journal, 2(2):92–110. find
  49. Andrew Williams. 2024. Delivering effective student feedback in higher education: An evaluation of the challenges and best practice. International Journal of Research in Education and Science, 10(2):473–501. find
  50. Helen Yannakoudakis, Ted Briscoe, and Ben Medlock. 2011. A new dataset and method for automatically grading ESOL texts. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, pages 180–189, Portland, Oregon, USA. Association for Computational Linguistics. find our card
  51. Jingheng Ye, Shang Qin, Yinghui Li, Xuxin Cheng, Libo Qin, Hai-Tao Zheng, Ying Shen, Peng Xing, Zishan Xu, Guo Cheng, and Wenhao Jiang. 2025. Excgec: a benchmark for edit-wise explainable chinese grammatical error correction. In Proceedings of the Thirty-Ninth AAAI Conference on Artificial Intelligence and Thirty-Seventh Conference on Innovative Applications of Artificial Intelligence and Fifteenth Symposium on Educational Advances in Artificial Intelligence, AAAI’25/IAAI’25/EAAI’25. AAAI Press. find
  52. Yue Zhang, Zhenghua Li, Zuyi Bao, Jiacheng Li, Bo Zhang, Chen Li, Fei Huang, and Min Zhang. 2022. MuCGEC: a multi-reference multi-source evaluation dataset for Chinese grammatical error correction. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 3118–3130, Seattle, United States. Association for Computational Linguistics. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/elhady-2026-arabigee-hierarchical-taxonomy-arabic-grammatical/