User community discovery from multi-relational networks
User community discovery from multi-relational networks
复制标题
DOI：
10.1016/j.dss.2012.09.012
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
Zhongfeng Zhang;Qiudan Li;D. Zeng;Heng Gao
中科院分区：
文献类型：
--
作者：
Zhongfeng Zhang;Qiudan Li;D. Zeng;Heng Gao
近年来，在线社交网络服务（SNS）经历了快速增长。 SNS使用户能够识别具有共同兴趣的其他用户、交换意见、建立交流论坛等。从社交网络中发现紧密连接的用户社区已成为主要挑战之一，以帮助理解 SNS 的结构属性并改进面向用户的服务，例如识别有影响力的用户和自动推荐。先前关于社区发现的工作将用户友谊网络和用户生成的内容分开处理。我们假设这两种类型的信息可以有效地整合，并提出一个在线社交网络中用户社区发现的统一框架。该框架将作者主题（AT）模型与用户友谊网络分析相结合。我们凭经验证明，这种方法能够使用两个真实世界的数据集发现有趣的用户社区。
Online social network services (SNS) have been experiencing rapid growth in recent years. SNS enable users to identify other users with common interests, exchange their opinions, and establish forums for communication, and so on. Discovering densely connected user communities from social networks has become one of the major challenges, to help understand the structural properties of SNS and improve user-oriented services such as identification of influential users and automated recommendations. Previous work on community discovery has treated user friendship networks and user-generated contents separately. We hypothesize that these two types of information can be fruitfully integrated and propose a unified framework for user community discovery in online social networks. This framework combines the author-topic (AT) model with user friendship network analysis. We empirically show that this approach is capable of discovering interesting user communities using two real-world datasets.