Vocational Prediction Method for High School Students Based on Machine Learning

Saravia Inga J.C.; Herrera Castillo V.R.; Garcia R.O.C. · Proceedings of the 2025 IEEE 32nd International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2025 (2025) · DOI: 10.1109/intercon67304.2025.11244629

Resumen

This article proposes a vocational prediction method for high school students based on Machine Learning (ML) techniques, specifically using Graph Neural Networks (GNN). Choosing a professional career is a crucial decision for students, but it is often influenced by subjective factors such as family pressure and friends' opinions, which can lead to a disconnect between the career choice and personal aspirations. This disconnect contributes to high dropout rates in higher education. The proposed model integrates psychometric profiles (RIASEC), academic performance, and socioeconomic data to recommend personalized career paths. By using GNNs, the model captures complex relationships between these factors, providing students with career suggestions tailored to their true interests and abilities. The system was evaluated using several GNN models, including GraphSAGE, GCN, and GAT, with GraphSAGE showing the highest accuracy in predictions. This approach aims to reduce dropout rates and help students make more informed vocational decisions aligned with their academic and professional future.

Instituciones: Universidad Peruana de Ciencias Aplicadas