Research On Collaborative Filtering Algorithm Based On User Interest

With the advent of the web 2.0 era,the amount of data is exponential growth.In the face of massive amounts of data,people are unable to quickly find the required resources.In order to solve the problem of resource retrieval and selection,scholars put forward the recommendation system.In the recommendation system,collaborative filtering algorithm is one of the most widely used recommendation algorithms nowadays.But in practical application,the traditional collaborative filtering algorithm does not take into account the user's interest and the item's own attributes in the recommendation process;thus affecting the quality of recommendation.In order to solve this problem,this paper mainly focuses on the two aspects of user interest and item attributes,and puts forward the relevant improvements and innovations as follows:1)Research and analyzes the present situation and deficiency of the existing recommendation algorithms.Through the user has evaluated the item attributes and ratings to establish user interest model;which can effectively solve the problem of user interest is not easy to capture.2)For the traditional collaborative filtering algorithm in the recommended process to ignore the recommended item's own attributes;therefore,a collaborative filtering algorithm based on user interest is proposed,the algorithm combines the attributes of the proposed item and the user interest model to give the user interest in the item,and the problem of neglecting the characteristics of the item itself in the recommendation process can be solved.3)Taking into account the deviation of user interest with time,the Ebbinghaus Forgetting Law and the sliding time-window are introduced to reflect the user's interest migration,and the user interest cooperative filtering algorithm is improved and optimized.The experimental data are selected from the classical MovieLens data set,and validates the proposed collaborative filtering algorithm based on user interest and the improved algorithm.The experimental results show that the proposed algorithm can effectively solve the problem of user interest and the cold start of the item in the recommendation process.The recommended quality is also improved.