Social Account Matching in Online Social Media using Cross-linked Posts

摘要

An online user creates his/her social account on multiple social media networks to interact with friends and relatives. Therefore, an individual produces diverse contents on each social network. As different social networks provide dissimilar but complementary services, integration of user-generated content collected from different social media profile of the same user may help us in understanding the actual behavior of the user. For this, we need to search the profile of a specific user on different social media services. In this paper, we propose a framework that exploits the user-generated posts to match the user identity across social networks. This framework uses a set of queries and automatically extracts unique usernames to build a dataset. Experimentally, we find that recall varies from 0.48 to 0.78 across different pairs of social networks.