An auto-recommending technology for 3G services

To improve the quality of 3G service recommending technology, a new 3G service recommendation algorithm was prompted based on clustering analysis and collaborative filtering. The algorithm can cluster the users with their behavior similarity to the commodities, and finds the nearest neighbor of an active user according to the clusters. Then the recommendation to the active user is produced by collaborative filtering. Experimental results show that the algorithm improves the performance of recommendation system and decreases the mean absolute error of 3G services system.