考虑时间情境的群体推荐算法研究

Group Recommendation Algorithm Research Considering Time Context

群体推荐能够有效地解决群体社交活动问题,存在着广阔的应用前景。然而现有的群体推荐研究是不完善的,很少考虑用户的时间情境信息。针对这个现状,在传统群体推荐方法的基础上考虑时间情境信息,利用时间衰减函数对评分进行预处理后再产生群体推荐列表。实验结果表明,所提出的算法能为群体提供有效的推荐,且引入时间情境可以提高推荐质量。

Group recommendation can effectively solve social activity problems that it has broad application prospects. How- ever, current group recommendation research is not sufficient that it rarely considers users＇time context information. In terms of this current situation, the proposed algorithm considers time context information on the basis of traditional group recommendation algorithm. Time decay function is used to preprocess ratings and then group recommendation list is generated. The experiemental results show that the proposed algorithm provide effective recommendation for groups and time context can improve recommenda- tion quality.