改进用户相似度的协同过滤算法应用研究
|
摘要
在基于用户的协同过滤推荐算法中,当用户-评分矩阵相对稀疏时,用户共同评分项目个数较少,皮尔逊相似度算法很难精确的计算用户之间的相似度,同时皮尔逊相似度算法对所有的商品赋予相同的相似度权重,没有考虑热门商品对相似度的影响。针对以上不足,文中在皮尔逊相似度算法的基础上,提出了一种改进的皮尔逊相似度公式,计算过程中考虑用户共同评价商品个数以及商品的热门程度这两个相似度影响因素,使得计算用户间相似度更加精确,从而获得更好的推荐效果。实验结果表明,文中改进的皮尔逊相似度算法能够在相似用户数较少时更加准确地计算用户之间的相似度,降低了平均绝对误差(Mean Absolute Error,MAE)。
In user-based collaborative filtering recommendation,when the rating matrix is relatively sparse,Pearson similarity algorithm is difficult to accurately calculate the similarity between users when the number of user common rating projects is small. The Pearson similarity algorithm gives out the same similarity weight to all goods without considering the influence of popular goods on similarity. To improve the above shortcomings,this paper considers two factors and comes up with an advanced Pearson similarity algorithm based on the traditional Pearson similarity algorithm. The experimental results show that the advanced Pearson similarity algorithm can calculate the similarity between users more accurately and reduced the mean average absolute error when the rating matrix is sparse.
In user-based collaborative filtering recommendation,when the rating matrix is relatively sparse,Pearson similarity algorithm is difficult to accurately calculate the similarity between users when the number of user common rating projects is small. The Pearson similarity algorithm gives out the same similarity weight to all goods without considering the influence of popular goods on similarity. To improve the above shortcomings,this paper considers two factors and comes up with an advanced Pearson similarity algorithm based on the traditional Pearson similarity algorithm. The experimental results show that the advanced Pearson similarity algorithm can calculate the similarity between users more accurately and reduced the mean average absolute error when the rating matrix is sparse.
引文
[1]王威,郑骏.基于用户相似度的协同过滤算法改进[J].华东师范大学学报:自然科学版,2016(3):60-66.
[2]孟庆庆,张胜男,卢楚雍.基于用户特征和商品特征的组合协同过滤算法[J].软件导刊,2015,14(3):41-43.
[3]郑翠翠,李林.协同过滤算法中的相似性度量方法研究[J].计算机工程与应用,2014,50(8):147-149,206.
[4]朱佳龙,吕玉琴,李巍海,等.一种新的基于非对称皮尔逊相似度的协同过滤算法[J].北京计算机科学技术,2014.
[5]Sarwar B,Karypis G,Konstan J,et al.Item-based collaborative filtering[J].Internet Computing,IEEE,2003,7(1):76-80.
[6]Resnick P,Iacovou N,Suchak M,et al.Group Lens:An open architecture for collaborative filtering of netnews[J].ACM,1994:175-186.
[7]Ahn H J.A new similarity measure for collaborative filtering to alleviate the new user cold-starting problem[J].Information Sciences,2008,187(1):37-51.
[8]Breese J S,Heckerman D,Kadie C.Empirical analysis of predictive algorithms for collaborative filtering[C]∥Fourteenth Conference on Uncertainty in Artificial Intelligence.Morgan Kaufmann Publishers Inc.1998:43-52.
[2]孟庆庆,张胜男,卢楚雍.基于用户特征和商品特征的组合协同过滤算法[J].软件导刊,2015,14(3):41-43.
[3]郑翠翠,李林.协同过滤算法中的相似性度量方法研究[J].计算机工程与应用,2014,50(8):147-149,206.
[4]朱佳龙,吕玉琴,李巍海,等.一种新的基于非对称皮尔逊相似度的协同过滤算法[J].北京计算机科学技术,2014.
[5]Sarwar B,Karypis G,Konstan J,et al.Item-based collaborative filtering[J].Internet Computing,IEEE,2003,7(1):76-80.
[6]Resnick P,Iacovou N,Suchak M,et al.Group Lens:An open architecture for collaborative filtering of netnews[J].ACM,1994:175-186.
[7]Ahn H J.A new similarity measure for collaborative filtering to alleviate the new user cold-starting problem[J].Information Sciences,2008,187(1):37-51.
[8]Breese J S,Heckerman D,Kadie C.Empirical analysis of predictive algorithms for collaborative filtering[C]∥Fourteenth Conference on Uncertainty in Artificial Intelligence.Morgan Kaufmann Publishers Inc.1998:43-52.