Background of the study
Multi-lingual academic libraries serve diverse user communities requiring access to resources in various languages. AI-driven translation tools—powered by neural machine translation—facilitate cross‑language information retrieval, interface localization, and user query translation (Khan, 2024). In higher education, such tools support international collaboration and broaden research access (Nguyen, 2025). Federal University, Dutsin‑Ma Library has implemented AI translation for its catalog interface, digital repository, and reference chat in English, Hausa, and Arabic (Eze, 2025). While preliminary usage indicates increased engagement among non‑English‑speaking patrons, systematic evaluation of translation accuracy, user satisfaction, and impact on resource discovery is lacking. Challenges include domain‑specific terminology, low-resource language pairs, and user trust in machine translations (Okoro, 2024). Evaluating AI translation tools within the library context will inform best practices for multilingual service provision in Nigerian academic settings.
Statement of the problem
Despite AI translation availability, non‑English‑speaking users report mistranslations of technical and legal education terms, leading to confusion and reduced resource utilization (Ibrahim, 2024). Without empirical assessment, the library cannot refine translation models or develop user training to ensure accurate multilingual access.
Objectives of the study
To assess the translation accuracy of AI tools for English–Hausa and English–Arabic educational content.
To measure user satisfaction and usage patterns of multilingual interfaces.
To identify model customization and user support needs for effective translation.
Research questions
What is the translation error rate for domain-specific library content?
How satisfied are users with multilingual library services?
What training and system improvements can enhance translation reliability?
Significance of the study
The evaluation will guide library management in selecting and fine‑tuning AI translation tools, designing multilingual interfaces, and developing training materials that empower all patrons to access academic resources effectively.
Scope and limitations of the study
This evaluation focuses on AI-driven translation in the catalog and digital repository of Federal University, Dutsin‑Ma Library. It excludes offline translation aids and non-library platforms. Limitations include evolving translation models and dialect variations.
Definitions of terms
Localization: Adapting interfaces and content to specific languages and cultural contexts.
Low-resource language: A language with limited digital corpora for AI model training.
Domain adaptation: Customizing AI models to specialized vocabulary and contexts.
Background of the Study
Library guides are curated online resources that provide step-by-step assistance on how to navigate...
Background of the Study:
Digital transformation has revolutionized library operations worldwide, reshaping the ways in whic...
Background of the Study
Radio has long been considered a key tool for promoting agricultural development, particularly in r...
Background of the study
Social media contests have emerged as innovative tools for enhancing brand engagement by activel...
Background of the Study
Extracurricular activities in early childhood education have gained prominence as an innovative ap...
Background of the Study
Public transport systems are vital for ensuring efficient urban mobility and stim...
Background of the Study
Consumer confidence is a vital barometer of economic health, reflecting the optimism or pessimism o...
Chapter One: Introduction
1.1 Background of the Study
Social media platforms have become a power...
Background of the Study
Urban migration in Nigeria has accelerated over the past decades, driving significant changes in la...
TECHNOLOGICAL ADVANCEMENT AND PSYCHO SOCIAL ISSUES AS DETERMINANT OF ACADEMIC PERFORMANCE
Abstract
...