Nigerian Researcher Explores Igbo Language Data for AI Development

A Nigerian linguistics researcher at the University of Chicago, Uduma U. Umeh, is generating data on Igbo as part of his research roadmap that could contribute to the development of Artificial Intelligence (AI) and language technologies for low-resource languages.

Umeh, a third-year PhD student, said his research in theoretical syntax seeks to explain how human language is structured and why languages exhibited particular structural patterns.

He said his research focuses primarily on Igbo, while also examining other West African languages, particularly those within the Benue-Congo language group.

According to him, his empirically driven research generates a rich database that could serve both linguistic scholarship and future computational applications.

“AI systems depend on linguistic data, and my research generates data that could eventually support language technologies for low-resource languages like Igbo,” he said.

Umeh said Igbo remains relatively under-resourced in Natural Language Processing (NLP) compared to English and other extensively studied languages.

He explained that his research on ellipsis, tone, predicate clefts and other grammatical structures document features that could eventually inform Igbo language technologies.

He said his work on Igbo ellipsis is particularly relevant because it examines cases where linguistic materials are absent, but their meaning could still be understood.

According to him, understanding such structures could help computational systems handle semantic interpretation, syntactic parsing, anaphora, machine translation and language generation.

“A system translating or interpreting Igbo needs to know when something that is not overtly expressed is nevertheless part of the intended structure,” he said.

Away from the AI perspective, there’s also the theoretical angle to Umeh’s work. He said that one of the core motivations for his research agenda is to properly situate Igbo and other Niger-Congo language substrate within broader universal theories of language.

According to him, this is an area that is largely lacking judging that only a few pockets of researchers are doing rigorous research on Igbo that engage contemporary theories of syntax.

“People mostly shy away from theoretical syntax because they consider it too abstract or even difficult. But someone must do syntax, however boring or mind-boggling it may seem to be,” he said.

“So, at the core of what I do is to contribute to ensuring that Igbo as well as other Niger-Congo languages have a place in the scheme of languages through whose data paradigm existing theories of syntax have been extended, applied, or even challenged,” he concluded.

And Beyond AI and the theoretical angle to all this, Umeh said his research could contribute to the preservation and teaching of Igbo, particularly among heritage speakers living outside Nigeria.

He explained that his contrastive studies of Igbo and English could identify areas where Igbo speakers might experience difficulties when learning or using English.

According to him, the findings could also provide a foundation for developing teaching materials for young people raised in Igbo-speaking families outside Nigeria.

He said such documentation is important because it preserves knowledge about the structure of Igbo for future generations while promoting cultural diversity.

Umeh holds master’s degrees in English from the University of Ibadan and Linguistics from the University of Chicago while still on his pathway to a PhD.

He said his academic journey has been demanding, requiring extensive research, writing and long hours of work.

His undergraduate research focused on error analysis, examining difficulties faced by secondary school students in Aba, Abia, in using English adjectives and adverbs.

He said the study identified errors including substitution, omission and overgeneralisation, which he attributed largely to negative transfer from students’ first languages.

Umeh said his progression into advanced linguistics research reflected his desire to move from consuming existing knowledge to producing new knowledge.

“I did not wake up one day and decide to venture into linguistics. It is more of a path I somehow found myself on,” he said.

“Linguistics chose me. I did not particularly choose it, but I am enjoying it.”

He said his broader objective is to situate Igbo within contemporary linguistics theories while producing knowledge that could benefit language education, preservation and emerging language technologies.