An Improved E-Learner Communities Self-Organizing Algorithm Based on Hebbian Learning Law

摘要

In this paper, we propose an improved E-Learner communities self-organizing algorithm based on Hebbian Learning Law, which can automatically group distributed e-learners with similar interests and make suitable recommendations to them. In particular, we build a collaborative agent for each e-learner to recommend resources and obtain ratings from neighbors, which results in trust weights updating and neighbors changing. Through similarity discovery, trust weights update, potential neighbors adjustment, reflexive adaption, and expert strategy, the algorithm presented in this paper implements an automatically adapted trust relationship with gradually enhanced satisfactions. It avoids difficult design work required for user preference representation or learner similarity calculation, which is suitable for open and distributed e-learning environments. A prototype has been built to simulate the self-organizing of the E-community and verify the validity of the algorithm. Experimental results have shown that the algorithm has preferable prediction accuracy and user satisfaction. In addition, we achieve an improvement on both satisfaction and scalability compared with the traditional Hebbian Learning algorithm proposed in earlier time.