【摘要】 为解决类别信息在推荐算法中因扁平或层次结构不能被充分利用的问题,提出一种基于组合类别空间的矩阵分解推荐算法。介绍项目组合类别空间,它是一种哈斯图结构,能更好组织类别信息;给出组合类别空间上的5种语义关系及距离度量,更好地描述用户偏好的动态变化;结合组合空间上的语义关系、距离及跳转次数等信息分别建立基于用户和基于项目的隐含特征矩阵模型。实验结果表明,提出方法的性能优于其它方法。更多还原

【Abstract】 To solve the problem that the category information cannot be fully utilized in the recommendation algorithm due to flat or hierarchical structure,matrix factorization recommendation algorithm based on combinational category space was proposed.The item combinational category space was introduced,which was a Hass diagram structure and could better organize category information.Five semantic relationships and distance metrics in the combinational category space were given,which better described the dynamic changes in user preferences.Implicit matrix feature models based on the user and item were established by combining semantic relations,distance and jump times in the combinational space.Experimental results show that the proposed method is better than other methods.更多还原