ISSN 1000-1239 CN 11-1777/TP

计算机研究与发展 ›› 2016, Vol. 53 ›› Issue (8): 1651-1663.doi: 10.7544/issn1000-1239.2016.20160202

所属专题: 2016数据挖掘前沿技术专题

• 人工智能 • 上一篇    下一篇



  1. (计算机软件新技术国家重点实验室(南京大学) 南京 210023) (江苏省软件新技术与产业化协同创新中心 南京 210023) (
  • 出版日期: 2016-08-01
  • 基金资助: 

A Ranking Based Poisson Matrix Factorization Model for Point-of-Interest Recommendation

Yu Yonghong, Gao Yang,Wang Hao   

  1. (State Key Laboratory for Novel Software Technology (Nanjing University), Nanjing 210023) (Collaborative Innovation Center of Novel Software Technology and Industrialization, Nanjing 210023)
  • Online: 2016-08-01

摘要: 随着基于位置社交网络(location-based social network, LBSN)的发展,兴趣点推荐成为满足用户个性化需求、减轻信息过载问题的重要手段.然而,已有的兴趣点推荐算法存在如下的问题:1)多数已有的兴趣点推荐算法简化用户签到频率数据,仅使用二进制值来表示用户是否访问一个兴趣点;2)基于矩阵分解的兴趣点推荐算法把签到频率数据和传统推荐系统中的评分数据等同看待,使用高斯分布模型建模用户的签到行为;3)忽视用户签到数据的隐式反馈属性.为解决以上问题,提出一个基于Ranking的泊松矩阵分解兴趣点推荐算法.首先,根据LBSN中用户的签到行为特点,利用泊松分布模型替代高斯分布模型建模用户在兴趣点上签到行为;然后采用BPR(Bayesian personalized ranking)标准优化泊松矩阵分解的损失函数,拟合用户在兴趣点对上的偏序关系;最后,利用包含地域影响力的正则化因子约束泊松矩阵分解的过程.在真实数据集上的实验结果表明:基于Ranking的泊松矩阵分解兴趣点推荐算法的性能优于传统的兴趣点推荐算法.

关键词: 基于位置社交网络, 兴趣点推荐, 泊松矩阵分解, BPR标准, 地域影响力

Abstract: With the rapid growth of location-based social network (LBSN), point-of-interest (POI) recommendation has become an important means to meet users’ personalized demands and alleviate the information overload problem. However, traditional POI recommendation algorithms have the following problems: 1)most of existing POI recommendation algorithms simplify users’ check-in frequencies at a location, i.e., regardless how many times a user checks-in a location, they only use binary values to indicate whether a user has visited a POI; 2)matrix factorization based POI recommendation algorithms totally treat users’ check-in frequencies as ratings in traditional recommender systems and model users’ check-in behaviors using the Gaussian distribution; 3)traditional POI recommendation algorithms ignore that users’ check-in feedback is implicit and only positive examples are observed in POI recommendation. In this paper, we propose a ranking based Poisson matrix factorization model for POI recommendation. Specifically, we first utilize the Poisson distribution instead of the Gaussian distribution to model users’ check-in behaviors. Then, we use the Bayesian personalized ranking metric to optimize the loss objective function of Poisson matrix factorization and fit the partial order of POIs. Finally, we leverage a regularized term encoding geographical influence to constrain the process of Poisson matrix factorization. Experimental results on real-world datasets show that our proposed approach outperforms traditional POI recommendation algorithms.

Key words: location-based social network (LBSN), point-of-interest (POI) recommendation, Poisson matrix factorization, Bayesian personalized ranking (BPR) metric, geographical influence