Research of the personalized recommender for E-Commerce based on web usage mining and collaborative filtering technique

Personalized recommender services of E-Commerce provides users with the preference items based on their Interest.Through web log mining,Forms the users´ access matrix,Calculate the similarity of users´ browsing habits and get the k-nearest neighbor users,According to neighbors´ project evaluation,forecast the target user´s evaluation of the project and give A top-N recommended items. Experiments show that the algorithm efficiency are achieved satisfactory recommendation results and solve the problem of new users in a degree.