Evaluación del desempeño basado en métricas de un sistema pedagógico multi-agente, ubicuo sensible al contexto y apoyado en ontologías

Abstract

The performance evaluation of a ubiquitous e-learning multi-agent system (U-MAS) is presented. The methodology is mainly based on four axes corresponding to ontological inferences, the context-awareness services, the mobile application and the overall performance of the U-MAS using qualitative and quantitative metrics. The results obtained suggest that the recommendations from ontological inferences are very successful, that the awareness services had great acceptance within the surveyed population of students, that the functionality concerning the mobile application was very satisfactory, and that the general performance evaluation of the U-SMA offers high levels of scalability for an increase number of concurrent users and with linear growth response. The results demonstrate the benefits of integrating ubiquitous computing, context-awareness services, adaptation to the student profile, as well as, a knowledge representation based on ontologies.