ENHANCING E-LEARNING COMMUNITIES THROUGH RECOMMENDER AGENTS (original) (raw)

The paper proposes a model for enhancing e-learning communities using recommender agents. The platform for creating such e-learning communities is described, as well as the provided recommendation mechanism. A recommender agent exploits web mining techniques and user's profile intensively to increase the members' satisfaction within an e-learning community. The recommender agent is presented in the context of a platform for self-built e-learning communities, in which a trainee enrols into the educational process, gains points through achievements and ultimately becomes a trainer, based on those points. The agent acts as a recommender system: it produces suggestions and shortcuts for learning materials, helping the learner to better navigate through online materials and quickly find needed resources. The main difference between suggestions and shortcuts, in the context of our recommender agents, is that shortcuts are created taking into account only the user's profile, but the suggestions are created using a collaborative recommendation mechanism. Besides recommending learning materials, our recommender agent can suggest suitable learning activities and tools (e.g. formative e-assessment, collaborative task management tools), depending on the optimal learning style of each user. In order to perform its functions, the recommender agent exploits an XML-based user profile, which can be filled either by the user himself or can be extracted from an e-learning system/social network, using customized connectors. The authors claim that the use of a recommender agent would personalize the educational experience of the e-learning community members, thus enhancing their engagement and increasing the efficiency of learning. In addition, the agent can be used as a stand-alone application, to create a personal learning environment, as well.