Artificial Intelligent-Powered Student Assistant for Modibbo Adama University, Yola

Student: Hamza Mai Abdulqadir (Project, 2025)
Department of Computer Science
Modibbo Adama University, Yola, Adamawa State

Abstract

ABSTRACT
This study presents the design, development, and evaluation of an AI-powered student assistant for Modibbo Adama University, Yola (MAU). The system was implemented as a mobile application using Java and Android Studio, with Firebase Firestore serving as the backend and Google’s Gemini API providing Natural Language Processing (NLP) capabilities. Decision tree algorithms were employed for academic risk prediction based on student records such as GPA and attendance, while k-means clustering was used to group students according to learning patterns and provide personalized recommendations. Testing and evaluation showed that the chatbot achieved 91% response accuracy, an average response time of 1.8 seconds, and a user satisfaction rating of 4.3/5. The system also successfully identified at-risk learners and provided tailored study strategies, demonstrating its effectiveness in enhancing access to academic information, reducing administrative workload, and improving personalized student support. By integrating localized institutional data with advanced NLP and machine learning techniques, this research contributes to knowledge on AI-driven educational tools in resource-limited contexts and establishes a framework for future deployment of intelligent academic support systems in Nigerian universities.

Keywords
Artificial Intelligence Chatbot Natural Language Processing Machine Learning Decision Tree K-Means Clustering Educational Technology Academic Support Student Performance Firebase
Filters
Institutions