Development of an AI-Based Student Depression Detection System Using the Multinomial Naïve Bayes Algorithm

Depression among students is a major mental health concern that interferes with both their emotional and academic. Depressed students go unnoticed since they hide their symptoms or confuse them with normal stress. Depression often affects focus and attention, making it hard for students to follow lessons, complete assignments, or retain information. Having a detection system in place enables early intervention that can help educators to detect warning signs early. This research used Agile Software Development Methodology in designing and developing a student depression detection system. This methodology enables continues feedback and enhancement to ensure the system is effective in meeting user needs. The process of development is segmented into short cycles referred to as sprints, with each sprint dedicated to developing particular features. For the evaluation, we selected eight (8) participants and five (5) IT experts for their technical expertise and three (3) mental health professionals for their relevance to the system’s purpose. ISO/IEC 25010 is a global standard that outlines software quality characteristics and evaluation criteria to ensure the systems' functionality, reliability, usability, efficiency, maintainability, and portability. With an overall weighted mean of 3.61 for the IT experts rated the system as highly acceptable. Similarly, the Mental Health Professionals evaluated the Student Depression Detection System with a highly acceptable rating of 3.71, reinforcing its support for student well-being. Researchers identified that the student depression detection system is a useful resource for the early identification of depression in students. It is advised to adopt the student depression detection system in colleges and schools to aid in the identification of vulnerable students.