[目的/意义]用户兴趣推荐是信息服务中的重要内容，针对目前融合情境信息推荐的研究更多是直接将情境作为单因素而缺乏考虑情境关联的思想，本文以情境关系为切入点实现社会化媒体用户的兴趣推荐。[方法/过程]以具有相似情境用户可能具有相似兴趣为假设，来改进用户原始兴趣网络从而实现推荐。通过社会网络和资源相似性计算构建原始兴趣网络中显性网络和隐性网络；借鉴共现原理和情境本身相似性构建情境网络；通过兴趣传递关系计算直接兴趣度与间接兴趣度；最后借鉴协同过滤的思想实现推荐。[结果/结论]与以往的只考虑单一情境因素的推荐方法相比，基于本方法的实验表明，将情境关系融入到推荐过程中不仅可以扩展用户的社会关系，而且可以得到更好的推荐效果。
[Purpose/significance] User interest recommendation is an important content in information services. The context of current researches about information recommendation is considered as an independent factor but ignoring the relation in it. This paper combined the contextual relations into the recommending process and realized the user interest recommendation on the social media. [Method/process] This paper takes a hypothesis that users in the similar context may have similar interests to improve the original user interest network and achieve the recommendation:constructing the explicit and implicit network by the social network and the resources similarity; constructing the contextual network by the co-occurrence principle and the context similarity itself; calculating direct interest scores and indirect interest scores by the network transmission; realizing recommendation by using collaborative filtering. [Result/conclusion] The experiment shows that combing contextual relations into the recommendation process can not only expand users' social relationship, but also ameliorate recommendation results.
[1] 袁园,孙宵凌,朱庆华. 微博用户关注兴趣的社会网络分析[J]. 现代图书情报技术,2012(2):68-75.
[2] JILIN G, WERNER G, CASEY D. "Make new friends, but keep the old"-recommending people on socialnetworking sites[C]//Proceedings of 27th annual CHI conference on human factors in computing systems,USA:HCI,2009:201-210.
[3] 郭磊,马军,陈竹敏,等. 一种结合推荐对象关联关系的社会化推荐算法[J].计算机学报2014,37(1):219-228.
[4] 姬朝阳. 一种基于标签耦合分析的微博用户关系挖掘方法[J]. 情报杂志,2015,34(2):164-167.
[5] SYEONIDIS P, NANOPOULOS A, MANOLOPOULOS Y. Tag recommendation base on tensor dimensionality reduction[C]//Proceedings of the 2008 ACM conference on recommender system, Switzerland:ACM,2008:43-50.
[6] DEY A K Providing architectural support for building context aware applications[D]. Atlanta:Georgia Institute of Technology,2000.
[7] JUN J J. Ubiquitous conference management system for mobile recommendation services based on mobilizing social networks:a case study of u-conference[J]. Expert systems with applications,2011,38(10):12786-12790.
[8] KARATZOGLOU A, AMATRIAIN X, BALTRUNAS L, et al. Multiverse recommendation:n-dimensional tensor factorization for context-aware collaborative filtering[C]//Proceedings of the fourth ACM conference on recommender system, Barcelona:ACM, 2010:79-86.
[9] 房小可. 一种融合情境因素的社会化信息推荐新方法[J]. 图书情报工作, 2015,59(22):105-129.
[10] 乔磊. 基于多维情境的一动信息服务个性化推荐算法研究[D].北京:北京交通大学,2016.
[11] 王栋,栾翠菊.基于情境信息的微博推荐算法研究[J].现代计算机,2016(6):33-38.
[12] 邢星. 社交网络个性化推荐方法研究[D].大连:大连海事大学,2013.
[13] BREESE J S, HECKERMAN D, KADIE C. Empirical analysis of predictive algorithms for collaborative filtering[C]//Proceedings of the fourteenth conference on uncertainty in artificial intelligence, Madison:UAI,1998.
[14] 易明,毛进,邓卫华.基于社会化标签网络的细粒度用户兴趣建模[J].现代图书情报技术,2011(4):35-41.
