Abstract:

The present proposal introduces an adaptive analysis system based on artificial intelligence (AI) designed to identify, characterize, and address learning difficulties among university students enrolled in the subject Materials Science and Metallurgical Engineering, taught across several engineering degrees.

The approach combines frequent formative assessments—through tests administered at the end of each thematic block—with an analytical model that categorizes student errors according to disciplinary content and the associated cognitive processing patterns (visual, verbal, numerical, spatial, or procedural). Based on the collected data, the system applies data mining techniques, clustering, and interpretable models to detect individual and group trends, distinguish types of difficulties, and generate learning profiles. These results enable instructors to plan and deliver targeted teaching sessions aimed at reinforcing the specific concepts and cognitive processing modes that present greater challenges for each student or group.

The goal is to provide early and personalized feedback that supports the balanced acquisition of the course competencies, improves the planning of sessions preceding the final evaluation, and promotes a more inclusive and evidence-based learning process. Preliminary results demonstrate the feasibility of this approach and its potential as a teaching support tool that can be integrated into any continuous assessment platform.

Keywords:

Material Science, Artificial Intelligence, Cognitive Patterns, Inclusive Learning.