Knowledge Recommendation Based On The Relative-Entropy Similarity Of Typicality

With the development of the Internet,the speed of information dissemination is accelerated.The knowledge resources has been becoming more and more various.Meanwhile,some online learning platform has been built.The new learning model is quite different from the old ones.The new learning model can makes the learners study at anywhere and anytime they want.Learners may be attracted by the plentiful knowledge resources on the Internet.Because everyone will find what they are interested in on the Internet.However,there are still some challenges when the knowledge resources on the Internet becoming much more people can deal with and the big data world comes.The recommendation system is a way to help people filter the information and find out the knowledge resources they are really interested in.But the knowledge resources are different from some other items,such as the music and films.So the existing recommendation system is not satisfied with the need of the online learners.The main contents in the research are:(1)Research on the process people learning on the Internet and set up the relationship between the users and the knowledge resources,using some idea from the developmental psychology.(2)Come up with the Typically-Based Relative Entropy Similarity recommendation method-TyRE.(3)TyRE is introduced into the knowledge recommendation.Set up the TyRE knowledge recommendation framework.In the research we first come up with the TyRE and introduce the TyRE into the knowledge recommendation.Meanwhile,we applied the framework on the online judge platform.The TyPE can help the learners study more efficiency on the Internet.The research will contribute to setting up a more intelligent,personalized online learning platform.