---
title: "SAHSOH@QALB-2015 Shared Task: A Rule-Based Correction"
description: "This paper describes the authors' entry in the QALB-2015 shared task on automatic correction of Arabic… 21 references, cited by 1 paper here."
authors: ["Wajdi Zaghouani", "Taha Zerrouki", "Amar Balla"]
year: 2015
venue: "Proceedings of the Second Workshop on Arabic Natural Language Processing"
source: "https://aclanthology.org/W15-3219/"
doi: "10.18653/v1/w15-3219"
licence: "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)"
licence_url: "https://creativecommons.org/licenses/by-nc-sa/3.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/zaghouani-2015-sahsoh-qalb-2015-shared-task/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zaghouani-2015-sahsoh-qalb-2015-shared-task/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2015](https://md.research.elamur.ai/free/en/papers/2015/page/3/) / SAHSOH@QALB-2015 Shared Task: A Rule-Based Correction Method of Common Arabic Native and Non-Native Speakers' Errors

# SAHSOH@QALB-2015 Shared Task: A Rule-Based Correction Method of Common Arabic Native and Non-Native Speakers' Errors

**Authors:** [Wajdi Zaghouani](https://md.research.elamur.ai/free/en/authors/wajdi-zaghouani/), Taha Zerrouki and Amar Balla · **Year:** 2015 · **Venue:** [Proceedings of the Second Workshop on Arabic Natural Language Processing](https://md.research.elamur.ai/free/en/venues/workshop-on-arabic-natural-language-processing/) ([Workshop on Arabic Natural Language Processing 2015](https://md.research.elamur.ai/free/en/venues/workshop-on-arabic-natural-language-processing/2015/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY-NC-SA 3.0 (ACL Anthology, before 2016)](https://creativecommons.org/licenses/by-nc-sa/3.0/)

**Links:** [Publisher's page](https://aclanthology.org/W15-3219/) · [PDF](https://aclanthology.org/W15-3219.pdf) · [DOI 10.18653/v1/w15-3219](https://doi.org/10.18653/v1/w15-3219) · [OpenAlex](https://openalex.org/W2251499398)

**Citations:** 7 citations (OpenAlex, 2026-10-14) · [1](https://md.research.elamur.ai/free/en/papers/zaghouani-2015-sahsoh-qalb-2015-shared-task/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
This paper describes the authors' entry in the QALB-2015 shared task on automatic correction of Arabic text. The task had two tracks: comments by native speakers on Aljazeera articles (L1) and essays by learners of Arabic (L2). The team chose not to use machine learning. They built a pipeline of hand-written rules instead. Regular expressions encode verb-form patterns such as infi'al and ifti'al together with their possible affixes. Replacement lists come from two lexicons, the Attia/AraComLex word list and JRC-Names. The authors injected typical letter confusions into these correct words and kept only the corrupted forms that are not valid words themselves. Further lists were tuned to each dataset using Hunspell with the Ayaspell dictionary. A manual error analysis guided the rules for each population. Punctuation was deliberately left untouched. On the 2015 test sets the system reached F1 53.97 (L1) and 25.12 (L2). When punctuation is excluded from scoring, these rise to 66.90 and 31.72. The work shows how far cheap, transparent rules can go and where they stop working.

## Key points
- Data: native track Alj-train/dev/test-2014, blind test Alj-test-2015; learner track L2-train/dev-2015, blind test L2-test-2015; scoring used the M2 scorer.
- Resources: Attia's list of about 9M inflected Arabic words (AraComLex), JRC-Names (1.18M person names and 6,700 organization names), the Alfaifi L1/L2 learner corpus, Hunspell with Ayaspell, and the authors' own Ghalatawi autocorrect tool (built on their earlier AkhtaBot for Arabic Wikipedia).
- Most frequent error types:
  - L1: Hamza, Ta-Marbuta/Ha and Alif-Maqsura/Ya confusions.
  - L2: definiteness, agreement, prepositions and word choice.
- Precision was high but recall was low, e.g. P 81.88 / R 40.24 on Alj-test-2015. The authors estimate that punctuation accounts for over 38% of the errors in QALB.
- Main failure cases:
  - split/merge errors;
  - typos that produce real words;
  - missing definite articles and gender/number agreement errors, which the authors say need a parser.

## Interesting ideas
- Build an autocorrect table by corrupting dictionary words with known confusions and discarding any corruption that is itself a valid word, so every entry maps to one unambiguous fix.
- Write rules over morphological templates plus affix slots instead of listing surface words, so one pattern covers a whole inflectional family.
- Generate correction lists separately for each corpus from spellchecker suggestions, dropping ambiguous cases to avoid needless replacements.
- Report scores both with and without a deliberately unhandled error class, so the real coverage of the system is visible.
- The authors note that real-word errors such as agreement and definiteness slip past word-level rules and need syntactic structure.

## For Mova

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

- **I1** (tokens): Apply this to English confusable spellings so the tokenizer can safely normalize noisy input without inventing ambiguous rewrites.
- **I2** (tokens): Express lemmatizer and morphology fixes as suffix/prefix templates, which keeps them few, auditable and explainable.
- **I3** (learn): Self-induced repair rules could be scoped per domain or genre and filtered by an ambiguity check before being trusted.
- **I4** (learn): Evaluations could routinely report results with known out-of-scope phenomena masked, as a control alongside the full score.
- **I5** (tree): Agreement checks on UD arcs (e.g. nsubj–verb number, amod–noun features) could flag likely input errors before frame building.

*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

- [Carnegie Mellon University](https://md.research.elamur.ai/free/en/universities/carnegie-mellon-university/) — University, Pittsburgh, United States · 539 works · 179,533 citations
- [Carnegie Mellon University Qatar](https://md.research.elamur.ai/free/en/universities/carnegie-mellon-university-qatar/) — University, Doha, Qatar · 47 works · 2,773 citations
- [École Nationale Supérieure d'Informatique](https://md.research.elamur.ai/free/en/universities/ecole-nationale-superieure-d-informatique/) — University, Algiers, Algeria · 8 works · 430 citations
- [University of Bouira](https://md.research.elamur.ai/free/en/universities/university-of-bouira/) — University, Bouïra, Algeria · 2 works · 26 citations

📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works · [🤝 Social Sciences](https://md.research.elamur.ai/free/en/shelves/social-sciences/) · 1,772 works · [🎓 Education](https://md.research.elamur.ai/free/en/shelves/social-sciences/education/) · 203 works

🧭 Topics (1): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 2,908 works

## Cite

Cite as: Wajdi Zaghouani, Taha Zerrouki and Amar Balla. 2015. SAHSOH@QALB-2015 Shared Task: A Rule-Based Correction Method of Common Arabic Native and Non-Native Speakers' Errors. Proceedings of the Second Workshop on Arabic Natural Language Processing.

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

```bibtex
@inproceedings{zaghouani2015sahsoh,
  title = {{SAHSOH@QALB-2015 Shared Task: A Rule-Based Correction Method of Common Arabic Native and Non-Native Speakers' Errors}},
  author = {Wajdi Zaghouani, Taha Zerrouki and Amar Balla},
  booktitle = {Proceedings of the Second Workshop on Arabic Natural Language Processing},
  year = {2015},
  doi = {10.18653/v1/w15-3219},
  url = {https://doi.org/10.18653/v1/w15-3219}
}
```

- [← previous Recognizing Textual Entailment using Dependency Analysis and Machine Learning](https://md.research.elamur.ai/free/en/papers/sharma-2015-recognizing-textual-entailment-dependency-analysis/)
- [next → TECHLIMED@QALB-Shared Task 2015: a hybrid Arabic Error Correction System](https://md.research.elamur.ai/free/en/papers/mostefa-2015-techlimed-qalb-shared-task-2015/)
- [↑ back to papers of 2015](https://md.research.elamur.ai/free/en/papers/2015/page/3/)

## References

21 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Alfaifi Abdullah and Atwell Eric. 2012. Arabic Learner Corpora (ALC): a taxonomy of coding errors. In Proceedings of the 8th International Computing Conference in Arabic (ICCA 2012) [↗](https://doi.org/10.13140/2.1.3126.8809)
2. Alfaifi, Abdullah and Atwell, Eric. 2013. Arabic Learner Corpus v1: A New Resource for Arabic Language Research. In proceedings of the Second Workshop on Arabic Corpus Linguistics (WACL-2). Lancaster University, UK. [find](https://scholar.google.com/scholar?q=Alfaifi%2C+Abdullah+and+Atwell%2C+Eric%2E+2013%2E+Arabic+Learner+Corpus+v1%3A+A+New+Resource+for+Arabic+Language+Research%2E+In+proceedings+of+the+Second+Workshop+on+Arabic+Corpus+Linguistics+%28WACL%2D2%29%2E+Lancaster+)
3. Alkanhal, Mohamed I., Mohamed A. Al-Badrashiny, Mansour M. Alghamdi, and Abdulaziz O. Al-Qabbany. 2012. Automatic Stochastic Arabic Spelling Correction With Emphasis on Space Insertions and Deletions. IEEE Transactions on Audio, Speech, and Language Processing, Vol. 20, No. 7, September 2012. [↗](https://doi.org/10.1109/tasl.2012.2197612)
4. Attia, Mohammed, Pavel Pecina, Younes Samih, Khaled Shaalan, Josef van Genabith. 2012. [Improved Spelling Error Detection and Correction for Arabic](https://md.research.elamur.ai/free/en/papers/attia-2012-improved-spelling-error-detection-correction/). COLING 2012, Bumbai, India.
5. Dahlmeier, Daniel and Ng, Hwee Tou. 2012. Better evaluation for grammatical error correction. In Proceedings of NAAC-HLT, Montreal, Canada. [find](https://scholar.google.com/scholar?q=Dahlmeier%2C+Daniel+and+Ng%2C+Hwee+Tou%2E+2012%2E+Better+evaluation+for+grammatical+error+correction%2E+In+Proceedings+of+NAAC%2DHLT%2C+Montreal%2C+Canada%2E)
6. Deorowicz S٫, Marcin G. Ciura. 2005. Correcting Spelling Errors By Modeling Their Causes. Int. J. Appl. Math. Comput. [find](https://scholar.google.com/scholar?q=Deorowicz+S%D9%AB%2C+Marcin+G%2E+Ciura%2E+2005%2E+Correcting+Spelling+Errors+By+Modeling+Their+Causes%2E+Int%2E+J%2E+Appl%2E+Math%2E+Comput%2E)
7. Golding and Roth. 1999. A Winnow based approach to Context-Sensitive Spelling Correction. In Machine Learning - Special issue on natural language learning, Volume 34 Issue 1-3, Feb. 1999. [↗](https://doi.org/10.1023/a:1007545901558)
8. Habash Nizar. 2010. Introduction to Arabic natural language processing. Synthesis Lectures on Human Language Technologies 3.1 (2010): 1-187 [↗](https://doi.org/10.1007/978-3-031-02139-8)
9. Habash Nizar, Ryan M. Roth. 2011. [Using Deep Morphology to Improve Automatic Error Detection in Arabic Handwriting Recognition](https://md.research.elamur.ai/free/en/papers/habash-2011-deep-morphology-improve-automatic-error/), ACL, page 875-884. The Association for Computer Linguistics, (2011)
10. Hadjir I . 2009 . Towards an open source Arabic spell checker. MA thesis in Natural language processing, scientific and technique research center to Arabic language development. [find](https://scholar.google.com/scholar?q=Hadjir+I+%2E+2009+%2E+Towards+an+open+source+Arabic+spell+checker%2E+MA+thesis+in+Natural+language+processing%2C+scientific+and+technique+research+center+to+Arabic+language+development%2E)
11. Hammad M and Mohamed Alhawari. 2010. In Recent improvement of arabic language search, Google Arabia Blog, Google company, 2010 [http://google-arabia.blogspot.com/](http://google-arabia.blogspot.com/). [↗](http://google-arabia.blogspot.com/)
12. Hassan Ahmed, Noeman Sara and Hassan Hany. 2008. Language Independent Text Correction using Finite State Automata. IJCNLP. Hyderabad, India. [find](https://scholar.google.com/scholar?q=Hassan+Ahmed%2C+Noeman+Sara+and+Hassan+Hany%2E+2008%2E+Language+Independent+Text+Correction+using+Finite+State+Automata%2E+IJCNLP%2E+Hyderabad%2C+India%2E)
13. Hirst Graeme and Alexander Budanitsky. 2005. Correcting real-word spelling errors by restoring lexical cohesion, Natural Language Engineering 11 (1): 87–111, 2005 Cambridge University Press Mohit Behrang, Alla Rozovskaya, Wajdi Zaghouani, Ossama Obeid, and Nizar Habash. 2014. The First shared Task on Automatic Text Correction for Arabic. In Proceedings of EMNLP Workshop on Arabic Natural Language Processing, Doha, Qatar. Obeid Ossama, Wajdi. Zaghouani, Behrang. [find](https://scholar.google.com/scholar?q=Hirst+Graeme+and+Alexander+Budanitsky%2E+2005%2E+Correcting+real%2Dword+spelling+errors+by+restoring+lexical+cohesion%2C+Natural+Language+Engineering+11+%281%29%3A+87%E2%80%93111%2C+2005+Cambridge+University+Press+Mohit+Behr)
14. Mohit, Nizar Habash, Kemal Oflazer and Nadi Tomeh. 2013. A Web-based Annotation Framework For Large-Scale Text Correction. In The Companion Volume of the Proceedings of IJCNLP 2013: System Demonstrations. Asian Federation of Natural Language Processing. [find](https://scholar.google.com/scholar?q=Mohit%2C+Nizar+Habash%2C+Kemal+Oflazer+and+Nadi+Tomeh%2E+2013%2E+A+Web%2Dbased+Annotation+Framework+For+Large%2DScale+Text+Correction%2E+In+The+Companion+Volume+of+the+Proceedings+of+IJCNLP+2013%3A+System+Demonstrati)
15. Pasha A., M. Al-Badrashiny, M. Diab, A. El Kholy, R. Eskander, N. Habash, M. Pooleery, O. Rambow, and R. Roth. 2014. [MADAMIRA: A Fast, Comprehensive Tool for Morphological Analysis and Disambiguation of Arabic](https://md.research.elamur.ai/free/en/papers/pasha-2014-madamira-fast-comprehensive-tool-morphological/). In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC).
16. Rozovskaya Alla, Houda Bouamor, Wajdi Zaghouani, Ossama Obeid, and Nizar Habash and Behrang Mohit. 2015. [The Second QALB Shared Task on Automatic Text Correction for Arabic](https://md.research.elamur.ai/free/en/papers/rozovskaya-2015-second-qalb-shared-task-automatic/). In Proceedings of ACL Workshop on Arabic Natural Language Processing, Beijing, China.
17. Shaalan, Khaled, Amin Allam and Abdallah Gomah. 2003. Towards automatic spell checking for Arabic. In Proceedings of the Conference on Language Engineering, 2003 - claes.sci.eg Steinberger, Ralf, Pouliquen, Bruno, Kabadjov, Mijail, Belyaeva, Jenya and van der Goot, Erik. 2011. JRC-NAMES: A Freely Available, Highly Multilingual Named Entity Resource. In Proceedings of the International Conference Recent Advances in Natural Language Processing, Hissar, Bulgaria. [find](https://scholar.google.com/scholar?q=Shaalan%2C+Khaled%2C+Amin+Allam+and+Abdallah+Gomah%2E+2003%2E+Towards+automatic+spell+checking+for+Arabic%2E+In+Proceedings+of+the+Conference+on+Language+Engineering%2C+2003+%2D+claes%2Esci%2Eeg+Steinberger%2C+Ralf%2C+Poul)
18. Zaghouani, Wajdi. 2014. Critical survey of the freely available Arabic corpora. In Proceedings of the Workshop on Free/Open-Source Arabic Corpora and Corpora Processing Tools Workshop, LREC 2014, Reykjavik, Iceland. [↗](https://doi.org/10.13140/rg.2.1.1362.1284)
19. Zaghouani Wajdi, Nizar Habash, Houda Bouamor, Alla Rozovskaya, Behrang Mohit, Abeer Heider, and Kemal Oflazer. 2015. Correction annotation for nonnative arabic texts: Guidelines and corpus. In Proceedings of The 9th Linguistic Annotation Workshop, pages 129–139, Denver, Colorado, USA, June. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Zaghouani+Wajdi%2C+Nizar+Habash%2C+Houda+Bouamor%2C+Alla+Rozovskaya%2C+Behrang+Mohit%2C+Abeer+Heider%2C+and+Kemal+Oflazer%2E+2015%2E+Correction+annotation+for+nonnative+arabic+texts%3A+Guidelines+and+corpus%2E+In+Proceed)
20. Zaghouani Wajdi, 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](https://md.research.elamur.ai/free/en/papers/zaghouani-2014-large-scale-arabic-error-annotation/). In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14), Reykjavik, Iceland. [↗](https://doi.org/10.13140/rg.2.1.3273.6169)
21. Zerrouki Taha. 2011. Improving the spell checking dictionary by users feedback. A meeting of experts check the spelling and grammar and composition automation, Higher Institute of Applied Science and Technology of Damascus, the Arab Organization for Education, Science and Culture, Damascus, April 18 to 20, 2011. [find](https://scholar.google.com/scholar?q=Zerrouki+Taha%2E+2011%2E+Improving+the+spell+checking+dictionary+by+users+feedback%2E+A+meeting+of+experts+check+the+spelling+and+grammar+and+composition+automation%2C+Higher+Institute+of+Applied+Science+and+)

*Updated 1 November 2026*