Collaborative Filtering Recommendation Algorithm Combining The User Background Information

As the internet and information technology rapidly develops,information overload made it hard for users to find the products;services;and so on they wanted within a mass of product information,especially in e-commerce.Therefore,the user would like personalized,targeted services.Many insiders are becoming concerned about how to provide more accurate individual information in less time to meet the actual needs for each user.Personalized information service includes personalized information search service,personalized information search and personalized information agent service.Recently,collaborative filtering recommendation systems have achieved widespread successes on the Web.However,the tremendous growth in the amount and variety of available information poses some austere challenges to recommendation systems,the problems of recommend accuracy,scalability,sparsity and cold-start are in dire need to be solved.Collaborative filtering requires a mass of the basal user data.Although in different applications,the data will be very different, but the current personalized recommendation system to provide personalized services kept most of the basic background information on the user.Driven by these facts,we make use of the existing user information to improve the traditional collaborative filtering.Therefore,we give an improved algorithm -- Collaborative Filtering Recommendation Algorithm combining the user background information.Firstly,the algorithm using the similarity of the attributes of items predicts the zero value and fills it in the user-item rating matrix,secondly,with the user's background information the algorithm computes the similarity between users,finally,the neighbor users are computed by user-based collaborative filtering recommendation and the final rate of forecast is presented.It not only can improve the accuracy of the user similarity but also increase the recommendation accuracy and the prediction accuracy of the final rating.We conduct a series of experiments to examine the effectiveness of our new algorithms;these experiments are all based on the MovieLens dataset.Comparing the MAE of the new algorithms to collaborative filtering recommendation algorithm based on item rating rrediction and traditional collaborative filtering algorithms,such as:User-based collaborative filtering algorithm,the experimental results show that new algorithms could effectively alleviate the data sparsity problem and reduce the MAE,and improv improve the recommendation accuracy efficiently.