Research On Collaborative Learning Recommendation System Based On Multi-Agent In E-learning

Nowadays, the e-learning characterized in multimedia technology, network technology and digital technology is becoming a kind of ideal learning way in the modern education. In the study of e-learning, the recommendation system and the collaborative learning has been the important research direction. The e-learning recommendation system can help learners locate suitable learning resources, so as to give learner the matched learning recommendation resources. And the collaborative learning integrates social factors into the e-learning, which can make learners cooperate in e-learning so as to effectively strengthen the mutual learning among learners. In addition, since agent technology has the advantages of intelligence and autonomy, integrating it into the e-learning system is a good way to enhance the working efficiency of the e-learning system and can better promote the learners' learning.Based on the above analysis, the paper proposes an idea which integrates intelligent agent technology into the e-learning environment combining collaborative learning with recommendation system, and establishes a framework model of this idea. The system can give learners at the same time the corresponding courseware recommendation service and collaborative learning service, so that learners can better learn the recommended courseware. Based on the learning courseware provided, a personalized courseware recommendation algorithm is proposed, which can dynamically estimate the learner ability based on the feedback of learner, and take into account the learner ability and the similarity between courseware texts to generate personalized courseware recommendation service for learners. And aimed at the collaborative learning provided, a personalized collaborative learning process is designed to eliminate the negative interaction among learners. At the same time, a personalized collaborative learning assessment and feedback method is proposed, which can analyze and evaluate the collaboration of learners based on the collaborative interaction of learners and the evaluation of team members, and provide learners with the corresponding visual feedback message.The experiment results show that the system built in the paper can indeed dynamically estimate the learner ability based on the learners' feedback and produce the matched learning recommendation coursewares for them. Meanwhile, it can also dynamically evaluate the collaboration of learners based on the collaborative interaction of learners and the evaluation of team members, and produce the understandable feedback information for learners, thus promoting their learning about the recommended courseware. In addition, the experiment results also indicate that compared with the separate individual learning, the collaborative learning indeed has the advantage in improving the learning effectiveness of learners.