复杂网络视角下跨社交网络用户身份识别研究综述  

Review of User Identification across Social Networks:The Complex Network Approach

作　　者：邢玲 邓凯凯 吴红海 谢萍 XING Ling;DENG Kai-kai;WU Hong-hai;XIE Ping(School of Information Engineering,Henan University of Science and Technology,Luoyang Henan 471023)

机构地区：[1]河南科技大学信息工程学院,河南洛阳471023

出　　处：《电子科技大学学报》2020年第6期905-917,共13页Journal of University of Electronic Science and Technology of China

基　　金：国家自然科学基金(61771185,61772175,61801171);河南省高校科技创新团队支持计划(21IRTSTHN015)。

摘　　要：社交网络是一种具有交互特性的复杂网络,利用复杂网络具有的网络特性可以链接不同社交网络中的节点,并分析节点之间存在的联系,结合相关的匹配算法可以有效地识别出用户在不同社交网络上的虚拟账号,有助于各大社交网络为用户提供更好的服务。该文对近十多年来数据挖掘领域中提出的跨社交网络用户身份识别技术进行了系统性地综述,详细阐述了3类用户身份识别技术相似度的计算方法和统一的识别框架,利用相关的评价指标对分类后的用户身份识别技术进行性能评估,最后展望了跨社交网络用户身份识别技术的未来研究方向。Social network is a complex network with interaction characteristics.It can link nodes in different social networks by using the network characteristics of complex network,analyze the connections between nodes,and combine with the related matching algorithm to identify user’s virtual accounts,which can help social networks to provide users with better services.This paper presents a systematic review on across social networks user identification techniques proposed in the field of data mining.Then the methods for calculating the similarity of the three types of user identification techniques and the unified identification framework are elaborated in detail.The relevant evaluation metrics are used to evaluate the classified user identification technique performances.Finally,the future research directions of across social networks user identification techniques are prospected based on the analysis of the research status.

关 键 词：跨社交网络 复杂网络 数据挖掘 实体用户 用户身份识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]