Developing Natural Language Processing Tools for Nigerian Languages: Opportunities, Challenges, and Future Directions

Abstract

This study explores the feasibility of developing Natural Language Processing (NLP) tools for Nigerian languages using the low-resource NLP framework (Joshi et al.) and the Language Resource Management Framework (LRMF) (Ndubuisi & Marivate) as analytical lenses. It aims to identify existing linguistic resources, assess the challenges hindering computational development, and propose strategies for enhancing technological adaptation within Nigeria's multilingual context. Using descriptive and exploratory research designs, data were synthesized from secondary sources, including NLP repositories such as Masakhane, PanLex, AI4D Africa, and published linguistic corpora for selected Nigerian languages. The synthesized evidence indicates that resource limitations, orthographic inconsistencies, fragmented language datasets, and inadequate institutional support remain major constraints to NLP development. However, the reviewed studies consistently identify community-driven corpus development, multilingual transfer learning, and collaborative open-source initiatives as promising approaches for improving computational support for under-resourced Nigerian languages. The study therefore recommends increased government investment in corpus development, greater integration of computational linguistics into higher education, and strengthened collaboration among researchers, language communities, and technology developers. By synthesizing current evidence, the study contributes to ongoing scholarship on language technology in Nigeria and Africa while informing future efforts toward inclusive and culturally responsive AI.