A Framework for Seed User Identification across Multiple Online Social Networks

With the advent of smart mobile phone social networking services are becoming the backbone of online virtual communication. Every year, several new social networks introduced with unique features that bring the opportunity for social network users' to migrate on the newly introduced social platform. In order to enjoy the services of newly introduced social media it is necessary that a user must have an account on that particular network. Finding the same user across the social network is a challenging task because of platform diversity. In this paper, we propose a hybrid personal information based approach to identify users across multiple online social networks. In this work, we utilize one content feature; 'cross link posts' and one network feature; 'following network' relation to identify unique seed users' across two social networks containing syntactically similar usernames namely, Twitter and Instagram. Matching user's 'following network' relationship using native approximate distance (Levenshtein) method, it is found that one cross link can give on an average 52 seed users. This result can be used as preprocessing step for large scale user identification across multiple social networks.