Survey reveals critical gaps in NLP research for Nigeria's major languages

Researchers have completed the first large-scale assessment of natural language processing efforts targeting Nigeria's three most widely spoken languages—Hausa, Yorùbá, and Igbo—which together reach over 175 million speakers. Despite this massive user base, the study reveals that these languages remain under-resourced for computational work.

Scale of the challenge#

Nigeria is home to more than 500 distinct languages, yet Hausa, Yorùbá, and Igbo account for roughly 65% of the linguistic landscape. However, the lack of sufficient digital materials and annotated datasets limits the development of modern NLP systems that could serve these populations. The dominance of English and other well-resourced languages in AI research means that speakers of these three major Nigerian languages often lack access to machine translation, voice assistants, and other language technologies that speakers of larger global languages take for granted.

What the review found#

The new survey, titled NaijaNLP, examined 293 studies focused on computational work across these three languages. A striking finding emerged: only about 28% of the reviewed research actually created or contributed new linguistic resources. This suggests the field has grown accustomed to recycling and repurposing the same existing datasets rather than systematically expanding the foundation of available training materials.

Unresolved technical problems#

The analysis identifies several language-specific hurdles that remain inadequately addressed. Handling morphological analysis—the study of how words break down into meaningful units—poses particular difficulty. Similarly, the proper representation and processing of diacritical marks, which carry semantic weight in languages like Yorùbá and Igbo, continues to be an under-explored area despite its practical importance for accurate language processing.

Path forward#

To strengthen NLP capabilities for low-resource languages like those in Nigeria, researchers call for strengthened collaboration among academic institutions, technology companies, and language communities. The priorities include expanding available training datasets through comprehensive annotation efforts, fostering sustained community involvement, and shifting the research culture toward resource creation rather than reliance on the same limited pools of existing data.