Classification Model Based on Chatbot and Unsupervised Algorithms to Determine Psychological Intervention Programs in Peruvian University Students

Huamán B.; Gómez D.; Lévano D.; Valles-Coral M.; Navarro-Cabrera J.R.; Pinedo L. · Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST (2022) · DOI: 10.1007/978-3-031-22324-2_15

Resumen

A strategy that supports the student’s academic and personal formation is that university consider tutoring as a mechanism that supports with favorable results to fight against the desertion of students. However, there are related problems in performing student segmentation and conducting psychological interventions. The objective was to formulate a classification model for intervention programs in university students based on unsupervised algorithms. For this, we carried out a non-experimental, simple descriptive study on a population of 60 university students; we carried out the data extraction process through a chatbot that applied the BarOn ICE test. After we obtained the data, the unsupervised k-means algorithm was used to group the students into sets determined based on the closest mean value obtained from the psychological test. We built a model for classifying students based on their answers to the BarOn ICE test based on K-means, with which we obtained five groups. The model classifies students by applying a different mathematical method to that used by the models applied by psychologists.

Instituciones: Universidad Nacional de San Martín; Universidad Peruana Unión