A New Approach for Recommender System

Abstract

In today's e-commerce environment, Collaborative Filtering (CF) is a widely used algorithm for recommender system, which is to identify the users who have similar preferences to the target user, and to predict the preference of the target user according to the preference ratings of the similar users. However, if the preference ratings of the target user are rare or none, then it cannot effectively identify the users with the similar preferences to the target user. In order to solve the problem of collaborative filtering, this study uses the implicit rating method to automatically calculate the user preference for the items by using the transaction data of the users, and further constructs an item-to-item, user-to-user, and user-to-item relationships, which can be used to calculate the preference rating for the target user, and recommend the products to the target user. The experimental results also show that the recommendation accuracy of our algorithm is higher than the other algorithms on average.