Research on collaborative filtering algorithm based on item''''s attribute implicit rating
基于商品属性隐性评分的协同过滤算法研究

After analyzing the previous work of recommendation system and algorithm, a recommendation algorithm based on Web daily record was presented which was accurate and efficient. Based on the behavior of customers, a CGA(Customer-Good-Attribute) model was designed which considered synthetically such factors as times, trails,commodities attributes and those distribution in Web pages.The function to the customers' implicit rating of the attributes of commodities was discussed,and a case was given. Then a filter algorithm was put forward based on attributes of goods. An experience result indicated that the algorithm is effective.