基于隐藏标签节点挖掘的跨网络用户身份识别

随着各种社交网络不断涌现，以及针对社交网络的安全和商业应用的不断普及，跨网络用户身份识别成为当前的研究热点。针对现有的基于自中心网络环境算法(Ego-UI)对标签节点利用率不高的缺点，该文提出一种基于隐藏标签节点挖掘的跨网络用户身份识别算法(HLNM-UI)。该算法通过给待匹配节点添加社团聚类信息，将挖掘出的隐藏标签节点加入到自中心网络里，通过对潜在的关系信息加以利用，提高待匹配节点的辨识度，然后利用标签节点找寻最佳匹配，最后通过迭代运算实现全网络所有节点的身份识别。在多个人工随机网络和真实社交网络实验结果表明，该文提出的算法相比现有的基于自中心网络算法具有更高的召回率和F-1值。

User identification across multiple networks based on hidden label nodes mining

With the emergence of various social networks and the expansion of security and business applications to social networks, user identification across multiple networks has been a popular research topic recently. For the problem that traditional algorithms based on Ego-network (Ego-UI) have low utilization of label nodes, this paper proposed a new algorithm of user identification across multiple networks based on Hidden Label Nodes Mining (HLNM-UI). By adding unmatched nodes with community clustering information, the proposed algorithm firstly mined hidden label nodes and took advantage of the potential relationship information to improve the identification degree of the nodes to be matched. Then it chose the highest similarity pair with label nodes as matching one. Finally, it conducted iterative operation to identify all nodes in the whole network. Experimental results on both synthetic random networks and online social networks show that the proposed algorithm has better recall and F-1 measurement than Ego-UI algorithms.