FUZZY LOGIC-BASED PROTOTYPE FOR PEER ASSESSMENT IN HIGHER EDUCATION

Artificial intelligence plays a fundamental role in educational research, especially in the assessment of problems involving uncertainty and imprecision. This article aims to model, within the peer assessment prototype, the integration of a computational model that correlates numerical and sentiment scores using fuzzy logic to generate individual and collective scores through measures of central tendency. The modelling method was applied and tested at the Technical University of Manabí (Ecuador) using the agile Scrum methodology. User testing was conducted to obtain feedback on the prototype's usability. The findings revealed that 80% of students and 90% of teachers did not experience any difficulties using the peer assessment prototype. Based on the recommendations provided by students and teachers, the necessary adjustments were made to the grade reporting interfaces. As future work, simulations are planned in massive scenarios to evaluate the prototype's performance and user satisfaction, as well as the implementation of cloud storage and processing solutions.