Research And Implementation Of Recommendation System Based On Interest Drift Collaborative Filtering Algorithm

With the rapid development of science and technology,the data in the Internet is entering the explosive growth.How to find the useful information needed by users in the huge data accurately and quickly has become the focus of research and recommendation.The two basic ways for users to obtain valuable information is the user's manual search and the system's personalized recommendation.In some areas,we need to recommend the user's desired information intelligently.The main contents of this paper are as follows:(1)Proposed the explicit and implicit interest drift detection models and drifted the interest.According to the phenomenon that the user 's interest changes with time,five interest models were proposed by explicit interest drift detection model,and the pattern matching was carried out by using the correlation attribute value and the density increment.We used the exponential forgetting function to carry out the implicit interest forgotten and then got the user's real interest.The experimental results showed that the explicit interest drift model achieved 88.23%~90.96% and 87.43%~89.54% accuracy and recall rate of users' interest,and can accurately capture the current interest of users.(2)Designed a collaborative filtering algorithm based on explicit and implicit interest drift detection model.Aiming at the problem of low recommendation accuracy of collaborative filtering algorithm,the similarity between attribute similarity and scale similarity was obtained;The project-based collaborative filtering algorithm was used to recommend the relevant items for the current user's interest.The experimental results showed that the average absolute deviation MAE value was reduced by 3.41% compared with the traditional cooperative filtering algorithm,which showed that compared with the traditional cooperative filtering algorithm,the detection of user interest and the calculation of mixed similarity can improve the recommendation accuracy rate of the algorithm.(3)Implementation of film recommendation system platform.Through the use of the key algorithms proposed in this paper,the film recommendation system was designed and implemented.The system use the SSH framework for background development and the platform provides users with the relevant recommendations of the film,as well as users to browse the project,to score and other functions.