Identification of Student Self-Blaming and Academic Burnout Based on Emotional Data Using the K-Nearest Neighbor Algorithm

Abstract

Academic demands and emotional pressure can lead students to experience self-blaming and academic burnout, which may negatively affect learning performance and psychological well-being. This study aims to identify student self-blaming and academic burnout based on emotional data using the K-Nearest Neighbor (KNN) algorithm. A quantitative research approach was employed using emotional data collected through a structured questionnaire measuring indicators such as feelings of guilt, emotional fatigue, stress, and decreased learning motivation. The collected data were processed through several stages, including data cleaning, normalization, and labeling, before being classified using the KNN algorithm. The results indicate that a considerable proportion of students experience moderate to high levels of self-blaming and academic burnout. The classification model achieved an accuracy of 75.15%, demonstrating satisfactory performance in identifying students’ emotional conditions. These findings confirm that emotional data can be effectively utilized as input features for machine learning–based classification in educational contexts. This study provides practical implications as an early detection mechanism to support data-driven academic guidance and psychological intervention programs. Academically, this research contributes to the integration of emotional data analysis, learning analytics, and machine learning techniques for student well-being assessment.