mlsanityData to 7 Oct 2026

NaijaNLP: A Survey of Nigerian Low-Resource Languages

arXiv 2502.19784 ↗First posted 27 Feb 2025cs.CL cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

With over 500 languages in Nigeria, three languages - Hausa, Yor`ub'a and Igbo spoken by more than 175 million people, account for about 65% of the languages. However, these languages are classed as low-resource due to insufficient digital resources to support tasks in computational linguistics. While several research efforts and initiatives have been presented, a coherent understanding of the state of classic Natural Language Processing (NLP) spanning grammatical formalisation to linguistic resources that support models development is lacking. This study presents the first comprehensive review of the state of affairs in NLP research across the three major Nigerian languages (NaijaNLP). We quantitatively assess the available linguistic resources and identify key challenges. Of the 293 reviewed studies, 27.6% contributed new linguistic resources. This finding highlights a strong reliance on repurposing existing data rather than creating new resources. Also, language-specific challenges, such as morphological analysis and effective representation of diacritics, remain under-explored. To advance NaijaNLP and LR-NLP more broadly, we echo the need for more collaborative efforts in resource enrichment, comprehensive annotation, and increased community support.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 54 of its 293 references that are in the map. First posted on arXiv 27 Feb 2025.

Builds on

54 of 293 references.

References · 54

  1. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  2. 2019Unsupervised Cross-lingual Representation Learning at Scalecited by 3,907
  3. 2020Don't Stop Pretraining: Adapt Language Models to Domains and Taskscited by 1,548
  4. 2018Datasheets for Datasetscited by 1,371
  5. 2014On the Properties of Neural Machine Translation: Encoder-Decoder Approachescited by 1,303
  6. 2022No Language Left Behind: Scaling Human-Centered Machine Translationcited by 1,290
  7. 2020The State and Fate of Linguistic Diversity and Inclusion in the NLP Worldcited by 789
  8. 2020Beyond English-Centric Multilingual Machine Translationcited by 559
  9. 2021The FLORES-101 Evaluation Benchmark for Low-Resource and Multilingual Machine Translationcited by 556
  10. 2020A Survey on Recent Approaches for Natural Language Processing in Low-Resource Scenarioscited by 170
  11. 2021MasakhaNER: Named Entity Recognition for African Languagescited by 155
  12. 2022Adapting Pre-trained Language Models to African Languages via Multilingual Adaptive Fine-Tuningcited by 149
38 more
  1. 2017Data Augmentation for Low-Resource Neural Machine Translationcited by 139
  2. 2020When Being Unseen from mBERT is just the Beginning: Handling New Languages With Multilingual Language Modelscited by 131
  3. 2021Survey of Low-Resource Machine Translationcited by 92
  4. 2020Low-resource Languages: A Review of Past Work and Future Challengescited by 90
  5. 2022NaijaSenti: A Nigerian Twitter Sentiment Corpus for Multilingual Sentiment Analysiscited by 64
  6. 2021Facebook AI WMT21 News Translation Task Submissioncited by 57
  7. 2022Towards Afrocentric NLP for African Languages: Where We Are and Where We Can Gocited by 46
  8. 2022AfroLM: A Self-Active Learning-based Multilingual Pretrained Language Model for 23 African Languagescited by 46
  9. 2024IrokoBench: A New Benchmark for African Languages in the Age of Large Language Modelscited by 43
  10. 2024Segment Any Text: A Universal Approach for Robust, Efficient and Adaptable Sentence Segmentationcited by 41
  11. 2022AfroLID: A Neural Language Identification Tool for African Languagescited by 38
  12. 2024Trans-Tokenization and Cross-lingual Vocabulary Transfers: Language Adaptation of LLMs for Low-Resource NLPcited by 35
  13. 2019Massive vs. Curated Word Embeddings for Low-Resourced Languages. The Case of Yor`ub'a and Twicited by 33
  14. 2017StarSpace: Embed All The Things!cited by 24
  15. 2023Noise-Robust Fine-Tuning of Pretrained Language Models via External Guidancecited by 24
  16. 2021OkwuGb'e: End-to-End Speech Recognition for Fon and Igbocited by 13
  17. 2022Hausa Visual Genome: A Dataset for Multi-Modal English to Hausa Machine Translationcited by 13
  18. 2024Correcting FLORES Evaluation Dataset for Four African Languagescited by 13
  19. 2023HaVQA: A Dataset for Visual Question Answering and Multimodal Research in Hausa Languagecited by 12
  20. 2024The Zeno's Paradox of `Low-Resource' Languagescited by 12
  21. 2018Attentive Sequence-to-Sequence Learning for Diacritic Restoration of Yor`ub'a Language Textcited by 11
  22. 2021NaijaNER : Comprehensive Named Entity Recognition for 5 Nigerian Languagescited by 10
  23. 2024NaijaHate: Evaluating Hate Speech Detection on Nigerian Twitter Using Representative Datacited by 9
  24. 2021The first large scale collection of diverse Hausa language datasetscited by 6
  25. 2019hauWE: Hausa Words Embedding for Natural Language Processingcited by 5
  26. 2020Improving Yor`ub'a Diacritic Restorationcited by 5
  27. 2020HausaMT v1.0: Towards English-Hausa Neural Machine Translationcited by 5
  28. 2021Translating the Unseen? Yoruba-English MT in Low-Resource, Morphologically-Unmarked Settingscited by 5
  29. 2021A Multilingual African Embedding for FAQ Chatbotscited by 5
  30. 2023NaijaRC: A Multi-choice Reading Comprehension Dataset for Nigerian Languagescited by 4
  31. 2023Detection and Analysis of Offensive Online Content in Hausa Languagecited by 4
  32. 2024Monolingual and Multilingual Misinformation Detection for Low-Resource Languages: A Comprehensive Surveycited by 4
  33. 2020Distant Supervision and Noisy Label Learning for Low Resource Named Entity Recognition: A Study on Hausa and Yor`ub'acited by 3
  34. 2020Development of a General Purpose Sentiment Lexicon for Igbo Languagecited by 3
  35. 2020The Challenge of Diacritics in Yoruba Embeddingscited by 3
  36. 2023Machine Learning Research Trends in Africa: A 30 Years Overview with Bibliometric Analysis Reviewcited by 3
  37. 2023HausaNLP at SemEval-2023 Task 12: Leveraging African Low Resource TweetData for Sentiment Analysiscited by 3
  38. 2024The State of Computer Vision Research in Africacited by 3

50 most cited of 54 All 54 (JSON)

Built on by

Not cited yet.