Research On The Entertainment Media Preference Matching Algorithm Based On The Difference Between Time Effect And Ratings

To improve the accuracy of users’ interest mining and achieve more accurate users’ personalized recommendations,an entertainment media preference matching algorithm based on the difference between time effect and ratings,named TRCF is presented.Firstly,in order to balance the differences in the impact of various factors such as users’ rating habits and product characteristics on the value of user-product ratings,the TRCF algorithm measures the difference in ratings between users based on the maximum difference in users’ ratings,and calculates the similarity of product preference among users based on the difference in ratings between users.Secondly,to capture the dynamic changes in users’ interest,the TRCF algorithm utilizes the number of times users interact with tags and the length of their lifecycle to mine users’ long-term and short-term interest,and combines the Ebbinghaus forgetting principle to calculate the similarity of tag preference between users.Thirdly,the product preference similarity and tag preference similarity between users are combined proportionally to obtain the similarity of comprehensive preference between users.Finally,based on the comprehensive preference similarity between users,find neighbors with similar interest to the target users and make recommendations.In order to check the recommended accuracy of the TRCF algorithm,the movie dataset movie Lens-25 M is used for the experimental verification of the TRCF algorithm.The experimental results show that the TRCF algorithm can effectively improve the recommendation accuracy of products.