Collaborative Filtering Recommendation Algorithm Based On Users' Feature And Preference

Collaborative filtering recommendation algorithm is one of the main algorithms of E-commerce recommendation system, and it is one of the most successful recommend technologies in the electronic commerce recommendation system. A collaborative filtering algorithm based on user features and preference is proposed. When calculate the users' similarity, join the user's features and preferences, First, calculated the users' features and preferences similarity based on the user's information, then calculate the users' similarity based on the users' co-ratings, and finally acquiring the last prediction result by combining the users'features and preference similarity and the similarity based on co-ratings with similarity weight, then recommend according to the prediction result..This article's experiment database is from the public MovieLens dataset, the experiment of the collaborative filtering recommendation algorithm was realized based on user features and background information, and comparing with the traditional collaborative filtering recommendation algorithm,. Experimental results show that the proposed based on user features collaborative filtering recommendation algorithm is superior to the traditional collaborative filtering recommendation algorithm based on the users. Through testing and analysis of the experiment data sets, the algorithm we proposed in the algorithm considered the users' feature and users'preference, increased the user similarity measure precision, and can consistently achieve better prediction accuracy, thereby brings better recommendation quality.