Elamur Research⌕
☰
home / papers / 2026 / Automatic Correction of Writing Anomalies in Hausa Texts

Automatic Correction of Writing Anomalies in Hausa Texts

Paper · 2026 · Ahmad Mustapha Wali and Sergiu Nisioi

Summary

Digital Hausa text often has two kinds of orthographic errors. The hooked letters ɓ, ɗ, ƙ and ƴ get replaced by plain Latin consonants, and spaces between words are dropped or added. Both can change a sentence's meaning and degrade NLP systems.

The authors combined Hausa Wikipedia, Wikimedia Hausa and other texts into more than 400,000 clean sentences. They then added synthetic noise calibrated against real errors in NaijaSenti tweets, which produced a parallel noisy–clean corpus. On this corpus they fine-tuned several seq2seq and African-language models, including M2M100, AfriTeVA, NCAIR1/N-ATLaS, UBC-NLP/cheetah-base, and BART and T5 variants.

The compact M2M100 (418M) matched 8B models fine-tuned for African languages. Correcting the text helped downstream classification, translation and LLM prompting, but did not fully restore the original quality.

The work matters because it gives Hausa, a low-resource language, its first transformer-based correction tools and an open dataset. The method could also be reused for other low-resource languages.

Key points

  • Error calibration: About 10,000 NaijaSenti tweets served as a seed for estimating realistic error patterns. Out-of-vocabulary tokens were clustered with DBSCAN (eps=0.4) over normalized Levenshtein distances, and the synthetic noise was tuned to match the resulting distribution.
  • Corpus composition: Hausa Wikipedia (160k sentences), Wikimedia Hausa (214k) and other texts (29k). About 10,000 words of MultiplEYE texts newly translated into Hausa serve as out-of-domain test data.
  • Downstream effect: Zero-shot tests with GPT-5.4 on noisy versus cleaned prompts showed the following:
    • MasakhaNER F1 went from 0.64 on original text to 0.19 when noised, recovering to 0.37 after cleaning. NER suffered most because word boundaries were not fully restored.
    • AfriXNLI scored 73 → 64 → 66 (original → noisy → cleaned).
    • AfriMGSM fell from 58 to 49 and did not recover after cleaning.
  • Stated limitations: The noise comes from social media but is projected onto edited texts. Synthetic errors may not suit every genre. The downstream evaluation relies on automatic metrics only.
  • Availability: The dataset is released under CC-BY-4.0, and the code is at github.com/ahmadmwali/HausaSeq2Seq.

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 BucharestBucharest · Romania21 works · 516 citations

 Fields, topics and concepts

📚 Shelves 5The 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 worksComputer Vision and Pattern Recognition · 3,486 works

Cite

Cite as Ahmad Mustapha Wali and Sergiu Nisioi. 2026. Automatic Correction of Writing Anomalies in Hausa Texts. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.430

@inproceedings{wali2026automatic,
  title = {{Automatic Correction of Writing Anomalies in Hausa Texts}},
  author = {Ahmad Mustapha Wali and Sergiu Nisioi},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.430},
  url = {https://doi.org/10.18653/v1/2026.acl-long.430}
}

References

60 references, 15 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Idris Abdulmumin, Sthembiso Mkhwanazi, Mahlatse S. Mbooi, Shamsuddeen Hassan Muhammad, Ibrahim Said Ahmad, Neo N. Putini, Miehleketo Mathebula, Matimba Shingange, Tajuddeen Gwadabe, and Vukosi Marivate. 2024. Correcting FLORES evaluation dataset for four African languages. In Proceedings of the Nint… find
  2. Amina Imam Abubakar, Abubakar Roko, AB Muhammad, and Ibrahim Saidu. 2019. Hausa wordnet: An electronic lexical resource. Saudi Journal of Engineering and Technology, 4(8):279–285. find
  3. Ife Adebara, AbdelRahim Elmadany, and Muhammad Abdul-Mageed. 2024. Cheetah: Natural language generation for 517 African languages. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 12798–12823, Bangkok, Thailand and virtual meet… find
  4. David Adelani, Jesujoba Alabi, Angela Fan, Julia Kreutzer, Xiaoyu Shen, Machel Reid, Dana Ruiter, Dietrich Klakow, Peter Nabende, Ernie Chang, Tajuddeen Gwadabe, Freshia Sackey, Bonaventure F. P. Dossou, Chris Emezue, Colin Leong, Michael Beukman, Shamsuddeen Muhammad, Guyo Jarso, Oreen Yousuf, and… — A Few Thousand Translations Go a Long Way! Leveraging Pre-trained Models for African News Translation
  5. David Ifeoluwa Adelani, Jade Abbott, Graham Neubig, Daniel D’souza, Julia Kreutzer, Constantine Lignos, Chester Palen-Michel, Happy Buzaaba, Shruti Rijhwani, Sebastian Ruder, Stephen Mayhew, Israel Abebe Azime, Shamsuddeen H. Muhammad, Chris Chinenye Emezue, Joyce Nakatumba-Nabende, Perez Ogayo, Are… — MasakhaNER: Named Entity Recognition for African Languages
  6. David Ifeoluwa Adelani, A. Seza Doğruöz, Iyanuoluwa Shode, and Anuoluwapo Aremu. 2025a. Does generative AI speak Nigerian-Pidgin?: Issues about representativeness and bias for multilingualism in LLMs. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 1571–1583, Albuquer… find
  7. David Ifeoluwa Adelani, Jessica Ojo, Israel Abebe Azime, Jian Yun Zhuang, Jesujoba Oluwadara Alabi, Xuanli He, Millicent Ochieng, Sara Hooker, Andiswa Bukula, En-Shiun Annie Lee, Chiamaka Ijeoma Chukwuneke, Happy Buzaaba, Blessing Kudzaishe Sibanda, Godson Koffi Kalipe, Jonathan Mukiibi, Salomon Kab… — IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models
  8. Ibrahim Ahmad, Shiran Dudy, Resmi Ramachandran-pillai, and Kenneth Church. 2024. Are generative language models multicultural? a study on Hausa culture and emotions using ChatGPT. In Proceedings of the 2nd Workshop on Cross-Cultural Considerations in NLP, pages 98–106, Bangkok, Thailand. Association… find
  9. Ibrahim Said Ahmad, Shiran Dudy, Tadesse Destaw Belay, Idris Abdulmumin, Seid Muhie Yimam, Shamsuddeen Hassan Muhammad, and Kenneth Church. 2025. Exploring cultural nuances in emotion perception across 15 african languages. Preprint, arXiv:2503.19642. find
  10. Adewale Akinfaderin. 2020. HausaMT v1.0: Towards English–Hausa neural machine translation. In Proceedings of the Fourth Widening Natural Language Processing Workshop, pages 144–147, Seattle, USA. Association for Computational Linguistics. find
  11. Jesujoba Oluwadara Alabi, Israel Abebe Azime, Miaoran Zhang, Cristina España-Bonet, Rachel Bawden, Dawei Zhu, David Ifeoluwa Adelani, Clement Oyeleke Odoje, Idris Akinade, Iffat Maab, Davis David, Shamsuddeen Hassan Muhammad, Neo Putini, David O. Ademuyiwa, Andrew Caines, and Dietrich Klakow. 2025.… — AFRIDOC-MT: Document-level MT Corpus for African Languages
  12. Axolotl maintainers and contributors. 2023. Axolotl: Open source llm post-training. find
  13. Muazzam Bashir, Azilawati Rozaimee, and Wan Malini Wan Isa. 2017. Automatic hausa language-text summarization based on feature extraction using naïve bayes model. World Applied Science Journal, 35(9):2074–2080. find
  14. Tadesse Destaw Belay, Israel Abebe Azime, Ibrahim Said Ahmad, David Ifeoluwa Adelani, Idris Abdulmumin, Abinew Ali Ayele, Shamsuddeen Hassan Muhammad, and Seid Muhie Yimam. 2025. AfroXLMR-social: Adapting pre-trained language models for African languages social media text. In Findings of the Associa… find
  15. Roger Blench. 2012. An atlas of Nigerian languages. Kay Williamson Educational Foundation Oxford. find
  16. Everlyn Asiko Chimoto, Constantine Lignos, Shamsuddeen Muhammad, Idris Abdulmumin, Clemencia Siro, and David Ifeoluwa Adelani, editors. 2026. Proceedings of the 7th Workshop on African Natural Language Processing (AfricaNLP 2026). Association for Computational Linguistics, Rabat, Morocco. find
  17. Berthold Crysmann. 2016. Representing morphological tone in a computational grammar of hausa. Journal of Language Modelling, 3(2). find
  18. Cheikh M Bamba Dione, David Ifeoluwa Adelani, Peter Nabende, Jesujoba Alabi, Thapelo Sindane, Happy Buzaaba, Shamsuddeen Hassan Muhammad, Chris Chinenye Emezue, Perez Ogayo, Anuoluwapo Aremu, and 1 others. 2023. Masakhapos: Part-of-speech tagging for typologically diverse african languages. In Proce… find
  19. David M. Eberhard, Gary F. Simons, and Charles D. Fennig. 2023. Ethnologue: Languages of the world. find
  20. Chris Emezue, Hellina Nigatu, Cynthia Thinwa, Helper Zhou, Shamsuddeen Muhammad, Lerato Louis, Idris Abdulmumin, Samuel Oyerinde, Benjamin Ajibade, Olanrewaju Samuel, Oviawe Joshua, Emeka Onwuegbuzia, Handel Emezue, Ifeoluwatayo A. Ige, Atnafu Lambebo Tonja, Chiamaka Chukwuneke, Bonaventure F. P. Do… — [arXiv:2304.12155](https://arxiv.org/abs/2304.12155) find
  21. Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, Naman Goyal, Tom Birch, Vitaliy Liptchinsky, Sergey Edunov, Edouard Grave, Michael Auli, and Armand Joulin. 2020. Beyond english-centric multiling… find
  22. Barry Haddow, Rachel Bawden, Antonio Valerio Miceli Barone, Jindřich Helcl, and Alexandra Birch. 2022. Survey of low-resource machine translation. Computational Linguistics, 48(3):673–732.
  23. Michael A. Hedderich, Lukas Lange, Heike Adel, Jannik Strötgen, and Dietrich Klakow. 2021. A survey on recent approaches for natural language processing in low-resource scenarios. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human…
  24. Nora Hollenstein, Marie-Luise Müller, Deborah N. Jakobi, Cui Ding, Maja Stegenwallner-Schütz, Ana Matić, Eva Pavlinušić Vilus, Ramunė Kasperė, Anna Bondar, Maroš Filip, Stefan Frank, Jana Hofmann, Thyra Krosness, Kaidi Lõo, Johanne Nedergaard, Chiara Tschirner, and Lena A. Jäger. 2026. MultiplEYE Da… find
  25. Kedir Yassin Hussen, Walelign Tewabe Sewunetie, Abinew Ali Ayele, Sukairaj Hafiz Imam, Shamsuddeen Hassan Muhammad, and Seid Muhie Yimam. 2025. The state of large language models for african languages: Progress and challenges. Preprint, arXiv:2506.02280. — [arXiv:2506.02280](https://arxiv.org/abs/2506.02280) find
  26. Aizhan Imankulova, Takayuki Sato, and Mamoru Komachi. 2019. Filtered pseudo-parallel corpus improves low-resource neural machine translation. ACM Transactions on Asian and Low-Resource Language Information Processing, 19(2):1–16. find
  27. Jesin James, Deepa P Gopinath, and 1 others. 2024. Advocating character error rate for multilingual asr evaluation. arXiv preprint arXiv:2410.07400. find
  28. Pratik Joshi, Sebastin Santy, Amar Budhiraja, Kalika Bali, and Monojit Choudhury. 2020. The state and fate of linguistic diversity and inclusion in the nlp world.
  29. Odunayo Jude Ogundepo, Akintunde Oladipo, Mofetoluwa Adeyemi, Kelechi Ogueji, and Jimmy Lin. 2022. AfriTeVA: Extending ?small data? pretraining approaches to sequence-to-sequence models. In Proceedings of the Third Workshop on Deep Learning for Low-Resource Natural Language Processing, pages 126–135… find
  30. Juraj Juraska, Daniel Deutsch, Mara Finkelstein, and Markus Freitag. 2024. MetricX-24: The Google submission to the WMT 2024 metrics shared task. In Proceedings of the Ninth Conference on Machine Translation, pages 492–504, Miami, Florida, USA. Association for Computational Linguistics.
  31. Alina Karakanta, Jon Dehdari, and Josef Van Genabith. 2017. Neural machine translation for low-resource languages without parallel corpora. Machine Translation, 32(1–2):167–189. find
  32. Ramunė Kasperė, Anna Bondar, Sergiu Nisioi, Maja Stegenwallner-Schütz, Hanne B. Søndergaard Knudsen, Ana Matić, Eva Pavlinušić Vilus, Dorota Klimek-Jankowska, Chiara Tschirner, Not Battesta Soliva, Deborah N. Jakobi, Cui Ding, Dima Abu Romi, Cengiz Acarturk, Matilda Agdler, Anton Marius Alexandru, M… find
  33. Tom Kocmi, Eleftherios Avramidis, Rachel Bawden, Ondřej Bojar, Konstantin Dranch, Anton Dvorkovich, Sergey Dukanov, Natalia Fedorova, Mark Fishel, Markus Freitag, Thamme Gowda, Roman Grundkiewicz, Barry Haddow, Marzena Karpinska, Philipp Koehn, Howard Lakougna, Jessica Lundin, Kenton Murray, Masaaki… — [arXiv:2508.14909](https://arxiv.org/abs/2508.14909) find
  34. Constantine Lignos, Idris Abdulmumin, and David Adelani, editors. 2025. Proceedings of the Sixth Workshop on African Natural Language Processing (AfricaNLP 2025). Association for Computational Linguistics, Vienna, Austria. find
  35. Stefano Lusito, Edoardo Ferrante, and Jean Maillard. 2023. Text normalization for low-resource languages: the case of ligurian. In Proceedings of the sixth workshop on the use of computational methods in the study of endangered languages, pages 98–103. find
  36. Shamsuddeen Hassan Muhammad, Idris Abdulmumin, Abinew Ali Ayele, Nedjma Ousidhoum, David Ifeoluwa Adelani, Seid Muhie Yimam, Ibrahim Sa’id Ahmad, Meriem Beloucif, Saif Mohammad, Sebastian Ruder, Oumaima Hourrane, Pavel Brazdil, Felermino D’ario M’ario Ant’onio Ali, Davis Davis, Salomey Osei, Bello S… — AfriSenti: A Twitter Sentiment Analysis Benchmark for African Languages
  37. Shamsuddeen Hassan Muhammad, David Ifeoluwa Adelani, Sebastian Ruder, Ibrahim Sa’id Ahmad, Idris Abdulmumin, Bello Shehu Bello, Monojit Choudhury, Chris Chinenye Emezue, Saheed Salahudeen Abdullahi, Anuoluwapo Aremu, Alípio Jorge, and Pavel Brazdil. 2022. NaijaSenti: A Nigerian Twitter sentiment cor…
  38. Shamsuddeen Hassan Muhammad, Ibrahim Said Ahmad, Idris Abdulmumin, Falalu Ibrahim Lawan, Sukairaj Hafiz Imam, Yusuf Aliyu, Sani Abdullahi Sani, Ali Usman Umar, Tajuddeen Gwadabe, Kenneth Church, and Vukosi Marivate. 2025. HausaNLP: Current status, challenges and future directions for Hausa natural l… find
  39. Wilhelmina Nekoto, Vukosi Marivate, Tshinondiwa Matsila, Timi Fasubaa, Tajudeen Kolawole, Taiwo Fagbohungbe, Solomon Oluwole Akinola, Shamsuddeen Hassan Muhammad, Salomon Kabongo, Salomey Osei, Sackey Freshia, Rubungo Andre Niyongabo, Ricky Macharm, Perez Ogayo, Orevaoghene Ahia, Musie Meressa, Mofe… — Participatory Research for Low-resourced Machine Translation: A Case Study in African Languages
  40. Paul Newman. 2000. The Hausa Language: An Encyclopedic Reference Grammar. Yale University Press, New Haven. find
  41. S. Nirenburg. 2009. Language Engineering for Lesser-Studied Languages. IOS Press. find
  42. Sergiu Nisioi, Anna Bondar, Ramunė Kasperė, and Maja Stegenwallner-Schütz. 2026. The multipleye text corpus data and materials. find
  43. A. NithyaKalyani and S. Jothilakshmi. 2019. Speech Summarization for Tamil Language, pages 113–138. Academic Press. find
  44. NLLB Team, Marta R. Costa-jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, Anna Sun, Skyler Wang, Guillaume Wenzek, Al Youngblood, Bapi Akula, Loic Barrault, Gabriel Mejia Gonzalez, Prangthip Hansanti, and 20 o… find
  45. Kelechi Ogueji, Yuxin Zhu, and Jimmy Lin. 2021. Small data? no problem! exploring the viability of pretrained multilingual language models for low-resourced languages. In Proceedings of the 1st Workshop on Multilingual Representation Learning, pages 116–126, Punta Cana, Dominican Republic. Associati…
  46. Arturo Oncevay, Gerardo Cardoso, Carlo Alva, César Lara Ávila, Jovita Vásquez Balarezo, Saúl Escobar Rodríguez, Delio Siticonatzi Camaiteri, Esaú Zumaeta Rojas, Didier López Francis, Juan López Bautista, Nimia Acho Rios, Remigio Zapata Cesareo, Héctor Erasmo Gómez Montoya, and Roberto Zariquiey. 202… find
  47. Shantipriya Parida, Idris Abdulmumin, Shamsuddeen Hassan Muhammad, Aneesh Bose, Guneet Singh Kohli, Ibrahim Said Ahmad, Ketan Kotwal, Sayan Deb Sarkar, Ondřej Bojar, and Habeebah Kakudi. 2023. HaVQA: A dataset for visual question answering and multimodal research in Hausa language. In Findings of th… find
  48. Matt Post. 2018. A call for clarity in reporting BLEU scores. In Proceedings of the Third Conference on Machine Translation: Research Papers, pages 186–191, Belgium, Brussels. Association for Computational Linguistics.
  49. Ochilbek Rakhmanov and Tim Schlippe. 2022. Sentiment analysis for Hausa: Classifying students’ comments. In Proceedings of the 1st Annual Meeting of the ELRA/ISCA Special Interest Group on Under-Resourced Languages, pages 98–105, Marseille, France. European Language Resources Association. find
  50. Saheed Abdullahi Salahudeen, Falalu Ibrahim Lawan, Ahmad Wali, Amina Abubakar Imam, Aliyu Rabiu Shuaibu, Aliyu Yusuf, Nur Bala Rabiu, Musa Bello, Shamsuddeen Umaru Adamu, and Saminu Mohammad Aliyu. 2023. Hausanlp at semeval-2023 task 12: Leveraging african low resource tweetdata for sentiment analys… find
  51. Sani Abdullahi Sani, Shamsuddeen Hassan Muhammad, and Devon Jarvis. 2025. Investigating the impact of language-adaptive fine-tuning on sentiment analysis in Hausa language using AfriBERTa. In Proceedings of the First Workshop on Language Models for Low-Resource Languages, pages 101–111, Abu Dhabi, U… find
  52. Iyanuoluwa Shode, David Ifeoluwa Adelani, Jing Peng, and Anna Feldman. 2023. Nollysenti: Leveraging transfer learning and machine translation for nigerian movie sentiment classification. arXiv (Cornell University). find
  53. Yuqing Tang, Chau Tran, Xian Li, Peng-Jen Chen, Naman Goyal, Vishrav Chaudhary, Jiatao Gu, and Angela Fan. 2020. Multilingual translation with extensible multilingual pretraining and finetuning. find
  54. Awarri Technologies, National Information Technology, and Development Agency. 2025. N-atlas-llm: A multilingual african language model. Fine-tuned Llama-3 8B model for African languages developed in collaboration with the Federal Government of Nigeria. find
  55. Kosei Uemura, Miaoran Zhang, and David Ifeoluwa Adelani. 2026. AfriMTEB and AfriE5: Benchmarking and adapting text embedding models for African languages. In Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 369…
  56. Aghaegbuna Uwaezuoke and Gift Anachunam. 2023. A contrastive study of reduplication in the igbo and hausa languages. Journal of The Linguistic Association of Nigeria, 26(1):82–105. find
  57. Adam Wiemerslage, Miikka Silfverberg, Changbing Yang, Arya McCarthy, Garrett Nicolai, Eliana Colunga, and Katharina Kann. 2022. Morphological processing of low-resource languages: Where we are and what’s next. Findings of the Association for Computational Linguistics: ACL 2022, pages 988–1007. find
  58. Curtis Wigington, Seth Stewart, Brian Davis, Bill Barrett, Brian Price, and Scott Cohen. 2017. Data augmentation for recognition of handwritten words and lines using a cnn-lstm network. In 2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR), pages 639–645. find
  59. Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, and Colin Raffel. 2021. mT5: A massively multilingual pre-trained text-to-text transformer. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Ling…
  60. Eric Zhu, Vadim Markovtsev, Aleksey Astafiev, Arham Khan, Chris Ha, Wojciech Łukasiewicz, Adam Foster, Sinusoidal36, Spandan Thakur, Stefano Ortolani, Titusz, Vojtech Letal, Zac Bentley, fpug, hguhlich, long2ice, oisincar, Ron Assa, Senad Ibraimoski, and 8 others. 2024. ekzhu/datasketch: v1.6.5. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wali-2026-automatic-correction-writing-anomalies-hausa/