The Design And Implementation Of Personalized Recommendation System Based On Video Knowledge Base

With the emergence and popularity of the Internet,what followed was a flood of data.However,in the face of a lot of information,the efficiency of the information reduced,this is so-called information overload problem.Well,personalized recommendation system is the response to the needs of the emergence of demand,It is based on users active or passive display of information needs,hobbies,historical behavior etc.obtained the training model through Large amount of data.With the combination of recommendation engine and video knowledge base,Video knowledge base provides flood of information and recommendation engine is a display vector Repository Information,convert knowledge base of information into content which user can use.Provide more extensive information rich text,Able to guide the user in a shorter time to obtain more information.This paper research and analysis the combination technology of personalized recommendation system and video knowledge base,design the personalized recommendation system based on video knowledge base.Including the following contents:design the architecture of personalized recommendation system based on video knowledge base.apply the features and video relationship provided by knowledge base to the recommended system.Including the feature information provided by the video knowledge base and the video features to establish the relationship between the use of video feature based on the similarity of the recommendation algorithm and video based recommendation algorithm.use collaborative filtering algorithm personalized recommendation,combining the three different personalized recommendation results using a variable weight distribution.Finally,this paper implementation,test and evaluation the design system.The recommended results of the proposed algorithm and the classical collaborative filtering algorithm are compared and the recommendation results under different weight distribution are compared.Verify that the personalized recommendation system based on video knowledge base can effectively solve the problem of user data sparsity,improve the quality of the recommended redults.