摘要
当前最新的兴趣点推荐工作开始融合地理、文本和社交信息进行推荐,但是还存在信息挖掘不充分的情况。为此,提出了改进的多类型信息融合的联合概率生成的兴趣点推荐模型。首先提出了自动学习文档话题数目的分层狄利克雷过程主题模型,学习用户和兴趣点相关兴趣话题;同时,利用由签到分布决定带宽大小的核密度估计法,个性化地理信息对用户签到行为的影响,而且还融合了用户位置访问序列中已访问兴趣点对待访问兴趣点的影响,即序列模式的影响;然后综合考虑了用户社交关系的影响;最后基于联合概率生成模型,融合文本、地理、社会和序列信息,提出TGSS-PGM兴趣点推荐模型,依据计算结果从而生成兴趣点推荐列表推荐给用户。实验结果表明,该模型在推荐准确率等多种评价指标上都取得了更好的结果。
The state-of-the-art studies started paying attention to comprehensively analyze geographical information,comment information and social information,but there is still insufficient information mining. To this end,this paper proposed a joint probabilistic generative model of multi-information fusion. Firstly,the framework learned interest topics of users and POIs through textual information,by exploiting the aggregated hierarchical Dirichlet process model,which could automatically learn the number of topics,to replace latent Dirichlet allocation model. Secondly,according to the kernel density estimation method,whose bandwidth depended on the check-in distribution,the framework conducted personalized modeling of geographic information. Thirdly,it also took consideration of sequential patterns,which was the impact of the visited location to the non-visited location. Then,it modeled social relevance comprehensively. At last,based on the joint probabilistic generative model,this paper proposed the TGSS-PGM model,exploiting multi-type contextual information and incorporating these factors effectively. Experimental results in real world social network show that the proposed model outperforms state-of-the-art recommendation algorithms in terms of precision and rating error.
作者
胡德敏
杨晨
Hu Demin;Yang Chen(School of Optical-Electrical & Computer Engineering;Institute of Computer Software & Technology,University of Shanghai for Science & Technology,Shanghai 200093,China)
出处
《计算机应用研究》
CSCD
北大核心
2018年第6期1636-1640,1675,共6页
Application Research of Computers
基金
国家自然科学基金资助项目(61170277
61472256)
上海市教委科研创新重点项目(12zz137)
上海市一流学科建设项目(S1201YLXK)
关键词
基于位置的社交网络
兴趣点推荐
隐马尔可夫链
核密度估计
话题模型
社交影响
location-based social network(LBSN)
location recommendations
additive Markov chain
kernel density estimation
topic model
social influence