<![CDATA[Natural Language Processing (NLP) plays a pivotal role in sentiment analysis, particularly in multilingual societies such as Nigeria, where languages like Igbo, Hausa, Yoruba, and Pidgin English coexist yet remain largely underrepresented in existing NLP tools and resources. This study aims to develop a sentiment analysis framework specifically tailored to Nigeria’s four major languages, addressing key challenges including code-mixing, tonal variations, and the scarcity of annotated datasets. The research leverages existing linguistic studies on these languages while constructing customized annotated datasets and designing models optimized for their structural and phonological properties. By systematically integrating linguistic insights with task-specific model development for low-resource settings, the proposed framework is designed to handle multilingual and code-mixed inputs more effectively than generic NLP systems. The study concludes that a targeted, language-aware approach is essential for improving sentiment analysis performance in underrepresented African languages and for ensuring that NLP technologies reflect the linguistic realities of multilingual societies. The contributions of this research lie in advancing multilingual sentiment analysis for low-resource African languages, providing methodological guidance for handling code-mixing and tonal features, and supporting practical applications in business intelligence, governance, public opinion mining, and social media analytics.]]>
Copyrights © 2026
