计算机工程与应用 ›› 2020, Vol. 56 ›› Issue (11): 67-74.DOI: 10.3778/j.issn.1002-8331.1901-0435

• 大数据与云计算 • 上一篇    下一篇

基于LBSN动态异构网络的时间感知兴趣点推荐

李全,许新华,刘兴红,林松   

  1. 湖北师范大学 教育信息与技术学院,湖北 黄石 435002
  • 出版日期:2020-06-01 发布日期:2020-06-01

Time-Aware Point-of-Interest Recommendation Based on Dynamic Heterogeneous Network in LBSN

LI Quan, XU Xinhua, LIU Xinghong, LIN Song   

  1. School of Educational Information and Technology, Hubei Normal University, Huangshi, Hubei 435002, China
  • Online:2020-06-01 Published:2020-06-01

摘要:

随着基于位置社交网络(Location-Based Social Network,LBSN)的快速发展,兴趣点(Point-Of-Interest,POI)推荐可以帮助人们发现有趣的并吸引人的位置。针对签到数据的稀疏性和用户兴趣的动态性等挑战性问题,提出了基于LBSN动态异构网络的时间感知兴趣点推荐算法。在LBSN异构网络模式中增加会话节点类型。通过动态元路径,在用户和兴趣点语义关系之间有效地融入时间信息、位置信息和社交信息等。设置了用户-兴趣点之间的动态元路径集,并提出了动态路径实例的偏好度计算方法。采用矩阵分解模型对不同动态偏好矩阵进行矩阵分解。根据不同动态元路径的用户特征矩阵和兴趣点特征矩阵,获取用户在目标时间访问兴趣点的推荐列表。实验结果表明,与其他兴趣点推荐方法相比,所提方法在兴趣点推荐精确度上取得了较好的推荐结果,具有良好的应用前景。

关键词: 基于位置社交网络, 兴趣点推荐, 动态异构网络, 时间感知, 矩阵分解

Abstract:

With the rapid development of Location-Based Social Network(LBSN), Point-Of-Interest(POI) recommendation can help people to discover interesting and attractive locations. Facing the challenging problems in LBSN, such as the sparseness of data and dynamics of user interest, it proposes the method of time-aware point-of-interest recommendation based on dynamic heterogeneous network in LBSN in this paper. Firstly, the session node type is added into the network schema in LBSN. The time information, position information and social information are integrated into the semantic relation between the user and POI by the dynamic meta-path. Secondly, the set of dynamic meta-path between user and POI is given, and the computational method of preference degree of dynamic path instances is proposed. Thirdly, the dynamic preference matrixes are factorized by the method of matrix factorization model. Finally, the POI recommendation list of user can be obtained at target time slot by the user feature matrixes and POI feature matrixs of different dynamic meta-paths. The experimental results show that the proposed method gets the best recommendation results in the aspect of precision of POI recommendation compared to other POI recommendation methods, and has a good application prospect.

Key words: location-based social network, Point-Of-Interest(POI) recommendation, dynamic heterogeneous network, time-aware, matrix factorization