User Clustering Slope One Algorithm Combined with Bhattacharyya Coefficient
WANG Wan-liang1，TU Hai-long1，ZHU Yan-liang1,ZHAO Yan-wei1，BAO Yi2
1（College of Computer Science and Technology，ZheJiang University of Technology，Hangzhou 310023，China）2（Hangzhou HeZhi Electronic Technology Co.，Ltd.，Hangzhou 310026，China）
Abstract：Slope One algorithm is a itembased collaborative filtering recommendation algorithm，which is simple and efficient，and the computational complexity is low.But the traditional Slope One algorithm treated the user’s rating data in a consistent manner，without considering the difference between the interests of users and difference between the similarity of items，affect the accuracy of the recommendation.Based on this，this paper improve it in the two dimensions of users and items，introducing the Bhattacharyya coefficient as the similarity between items，in the user dimension，and using clustering method to eliminate differences in user behavior.Finally，the experimental results show that the proposed method can guarantee the lower computational complexity，at two indicators，the MAE and the RMSE ，it have higher recommendation accuracy in the MovieLens data set.
王万良,屠海龙,朱炎亮,赵燕伟,鲍毅. 融合巴氏系数的用户聚类Slope One算法[J]. 小型微型计算机系统, 2018, 39(3): 539-543.
WANG Wan-liang,TU Hai-long,ZHU Yan-liang,ZHAO Yan-wei,BAO Yi. User Clustering Slope One Algorithm Combined with Bhattacharyya Coefficient. Journal of Chinese Computer Systems, 2018, 39(3): 539-543.
2018, Vol. 39 
