Facilitate collaborative recommendation based on community knowledge awareness tool

In large-scale e-learning environment, it is very difficult or even impossible for the tutors to give “one-to-one” instruction to the distributed learners. In this paper, we first presented a multi-agent framework to group e-learners into communities with similar learning status. Then through an integrated visualization platform, we enabled the tutors to gain insight to the characteristics of each individual community including knowledge structure, learning progress and existing problems. Based on this, the tutors can then assign further reading, group discussion to give personalized instructions to each community. Furthermore, the tutors can collaborate with the e-learners in the process of community construction to achieve better efficiency. Experimental results derived from real application have shown that this collaboration between the e-learners and tutors indeed improves the learners’ learning achievements. Key-Words: Collaborative Recommendation, E-Learning, Computer Supported Collaborative Learning