Model of a GPT-4 and Natural Language Processing Driven Chatbot for Mental Health Strategies Among University Students in Peru

Luna J.; Jara N.; Pozo K. · Proceedings of the LACCEI international Multi-conference for Engineering, Education and Technology (2026) · DOI: 10.18687/laccei2026.1.1.241

Resumen

Youth depression and suicide represent an increasingly critical public health challenge in Peru, with a significant impact on the university population. This paper presents a technological model based on an emotional-support chatbot that integrates the GPT-4 architecture and natural language processing (NLP) techniques. The approach combines sentiment analysis with the generation of empathetic responses, complemented by the delivery of educational resources. A quasi-experimental pretest-posttest design without a control group was conducted with a sample of 30 university students over a one-month intervention period. The evaluation included changes in depressive symptoms using the PHQ-9 scale, usability through the System Usability Scale (SUS), and perceived empathy and response performance. The results show an observed reduction in PHQ-9 scores, high usability, and positive user perception of empathy, suggesting that the chatbot provides a safe and accessible space for emotional support. However, these findings should be interpreted cautiously due to the limited sample size and absence of a control group. This proposal aims to strengthen preventive mental health strategies in the Peruvian university context through the use of innovative digital tools.

Instituciones: Universidad Peruana de Ciencias Aplicadas