The Research Of Personalized Recommendation Of Learning Resources Based On Collaborative Filtering Recommendation Technology

E-learning has become an effective way to carry out staff training. Currently, manylarge enterprises build e-learning platform in their corporate intranet or on the Internet.With the application of e-learning practice in-depth, the number of learning resources ine-learning platform is growing rapidly, resulting in that the learners need spend a lot oftime and effort to retrieve the platform and get learning resources to meet their ownneeds, or even cannot find learning resources they want. By analyzing numbers ofenterprises’ e-learning websites, we found there are several common ways of learningresources pushing in e-learning: such as Top-N, query with keywords. All of theabove ways can help learners find learning resources to some extent, to meet their needs,but cannot push personalized learning resources to learners. Therefore, how to pushlearning resources proactive and meet the needs of each learner becomes one of thethemes of educational technology researchers study.First, this paper analyzes and researches the collaborative filtering technology andalgorithms which has been successfully applied in e-commerce, then summarizes theshortcomings of the algorithm, and introduces several typical optimization algorithmsto address these issues.Secondly, the paper proposes a solution using the Content Filtering technology andRating Prediction Algorithm to solve the data sparse and cold-start problem of collaborative filtering algorithms. At the same time, build an implicit learning scoring model based onusers’ behavior. We obtain an optimized algorithm through integrating all of algorithmsabove.Finally, build a personalized recommendation model of learning resources basedon the optimized algorithm above. The model is presented through the knowledge basemodule of the information platform for small and medium enterprises in ZhejiangProvince.This study eases the sparse data and cold start problems of collaborative filteringalgorithms to a certain extent, but has not addressed it completely. By introducing thecollaborative filtering algorithms to the e-learning, we bring it into a new applicationarea. It will inspire more e-learning researchers to explore collaborative filteringtechnology in e-learning application from different levels and angles, and ultimatelyimprove the accuracy and efficiency of personalized recommendations which based on thecollaborative filtering algorithms.