Collaborative Filtering Algorithm Based On User Activity And Project Popularity

The problem of information overload is a difficult problem faced by today's mobile internet era.The rapid growth of data has led to a decrease in information utilization.It is difficult for users using traditional search tools to find items of interest.Personalized recommendation is the key technology to solve the problem of information overload.The core idea of the technology is to predict user preferences based on the user's historical behavior and purchase records,and to actively recommend items that users are interested in.Collaborative filtering algorithm is currently the most mature personalized recommendation algorithm,which is widely used in various fields.The algorithm predicts the user's most likely favorite things based on the user's existing project ratings.However,in the actual recommendation,the traditional collaborative filtering does not consider the impact of user activity and project popularity on the recommendation performance.Usually,the user's shopping behavior is closely related to the user's active and project popularity.For example,the more active users of the site,the more It is possible to browse the website's unpopular project,stating that these two factors will affect the accuracy of the recommendation system to predict user preferences,so the text proposes an improved algorithm that takes into account user activity and project popularity.This article combs the development background of personalization research and the current research status,introduces several most common recommendation algorithms,and focuses on the improvement ideas and implementation steps of collaborative filtering algorithms based on user activity and project popularity,and improves the algorithm.The active user's interest contribution value is reduced and the impact of popular products on the prediction score is weakened.Experiments show that the optimized collaborative filtering algorithm has high recommendation accuracy,a wider range of recommended coverage projects,and more practical significance in the field of e-commerce.In this paper,a combination of case studies and experimental studies is used to test and evaluate traditional collaborative filtering algorithms,algorithms that consider user activity and project popularity,and improved algorithms.The differences between algorithms and the specific optimization process are discussed..The simulation experiment was conducted through Python programming and the Movielen data set was used to test.The experimental results show that the improved algorithm presented in this paper shows better recommendation performance.