MBADIWE, CHINEMEREM (2026) DESIGN AND IMPLEMENTATION OF AN INTELLIGENT ACADEMIC SUPPORT AND PROGRESS TRACKING INFORMATION SYSTEM. Other thesis, Godfrey Okoye University, Enugu Nigeria.
![[img]](https://eprints.gouni.edu.ng/style/images/fileicons/text.png) |
Text
Mbadiwe_Chinemerem_Corrected_Project.pdf Download (1MB) |
Abstract
This project provides a solution to the common problem of periodical unavailability and lack of personalization in the academic system, especially in under-resourced areas. It does this by developing an intelligent academic support system that is always available. In many education systems, learning platforms sometimes lack personalized guidance for each individual student, because of this, the systems fail to identify struggling students until it is too late for teachers to provide meaningful help. The students may have already failed an exam before they are identified as “struggling”. To solve this problem, the project uses an approach where a machine learning model observes the activities of students on the learning platform and is trained to identify those who are lagging behind or at risk of failing very early in the semester. It uses a Random Forest machine learning model to observe and monitor the student’s activities closely. Using the information, the system alerts teachers early if a student is at risk of failing their courses. Furthermore, the platform includes an AI powered tutor to assist learners at all times. The AI tutor is powered by Google's Gemini AI, and a technology called “Retrieval-Augmented Generation” (RAG) is used to fine tune the AI tutor so it gives responses based on the learning materials of the school. The primary purpose of the AI tutor is to provide students with continuous, 24/7 access to academic support. The project’s software system was built using Flutter for the frontend and Flask framework for the backend. The choice of development framework helps create a simple and beautiful user interface and a backend that will run smoothly on the web (because Flask framework is very effective for web development). After development, the system was thoroughly tested to verify that its performance is good and that the system is reliable. The evaluation results were very promising for the progress tracking system. This solution showed the accuracy of 90.0%, 89.3% precision and 83.3% F1 Score. Such figures prove that the system is able to detect lagging students effectively and give almost no false alerts. The tests of the separate components of the system including the training sessions and live classroom confirmed all use cases to be applicable. The testing procedure also included the synchronization between the teacher's and student's dashboards ensuring that the updates on the latter would automatically update the former. Furthermore, the testing confirmed that the AI tutor consistently delivers reliable answers because the “RAG” technology ensures the answers are given based on the school’s materials. The system transforms academic support from a reactive system into a proactive process. However, future improvements are recommended to better support areas with slow internet connections.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Computing And Information Technology (FACIT) |
| Depositing User: | Cynthia Ugwuoti |
| Date Deposited: | 29 Jul 2026 13:05 |
| Last Modified: | 29 Jul 2026 13:05 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/6075 |
Actions (login required)
 |
View Item |
