基于用户关系的跨社交网络用户身份关联方法

为识别出不同社交网络平台中属于同一自然人的账号，提出了一种基于用户关系的跨社交网络用户身份关联方法。首先，设计了基于网络表示学习的用户关系提取模块，将大规模用户关系转换至低维向量空间进行表示；然后，针对异构信息网络改进了传统网络表示学习算法，提出了CSN_LINE算法，实现融合跨社交网络先验关联关系的网络表示；最后，构建了基于多层感知机的用户身份关联模型。实验结果表示，提出的方法与目前先进的方法相比，综合指标F1值和正确率的提高均超过12%，证明了该方法的合理性和有效性。

User identity linkage across social networks based on user relations

In order to distinguish the accounts that belong to the same person, this paper proposed a method to link user identity across social networks based on user relations. Firstly, it designed a user relations feature extraction module based on network representation learning. It could embed large information networks into low-dimensional vector spaces. Secondly, it proposed CSN_LINE algorithm for heterogeneous information network. The improved algorithm could represent network combining with anchor links across networks. Finally, this paper constructed a user identity linkage model based on multi-layer perception. Experiments show that the F1 rate and accuracy rate of this method increase over 12% compared with the current advanced algorithm, which prores the validity and rationality of the method.