Improving collaborative learning by personalization in Virtual Learning Environments using agents and competency-based ontology

With the spread of distance education courses or blended, an increasingly common problem is the lack of a personalized accompaniment to the student and the delay in responding by mediators and other colleagues the doubts and requests from students in Virtual Learning Environments (VLEs), usually posted on forums or manifested via e-mail. The approach taken to solving this problem presented in this paper is based on multi-agent systems and a competency-based ontology of the learner model. Through such technologies, the student's doubts are identified and these are directed to community members who have the profile with best suited skills and competencies to resolve it, decreasing the response time of doubts from students. A Petri Net has been developed in order to represent interactions among agents. The system was tested in a Numerical Analysis class that makes use of the VLE Moodle. The results of the tests were used to prove the validity of the system and the viability of the solution. Survey questionnaires were passed in the classroom in order to obtain and evaluate the students' impressions about the resource available to them in the VLE.