EDUCATIONAL CONTENT RECOMMENDATION PLATFORM USING LLMS AND FUZZY LOGIC

Abstract

This article sought to propose an educational platform that uses Large Language Models (LLMs) and Fuzzy Logic to recommend content in a personalized way. The methodology employed involves the use of LLMs for analysis, summarization of didactic materials, and keyword extraction, while Fuzzy Logic classifies the student's level based on their performance in gamification activities, such as QUIZ and Crosswords, considering parameters like score, attempts, difficulty, and time. The main results indicated that LLMs were effective in content summarization and keyword generation for searching relevant supplementary material. The Fuzzy system successfully classified student performance, allowing for adaptive content direction. It is concluded that the integration of these technologies offers dynamic and adaptive support for learning, personalizing the educational experience by providing content aligned with the student's needs and knowledge level, overcoming generic approaches and information overload.