Hybrid de-anonymization across real-world heterogeneous social networks

To enjoy various utility and services, people are active in multiple social networks nowadays. With tons of data generated on platforms, multiple accounts of the same user in different social networks can be used to de-anonymize the user in a large scale. The aggregation of user profiles poses a threat to user privacy. With a concern of privacy leakage, de-anonymization techniques, including graph based approaches and profile based approaches, are widely studied in recent years. However, few works throw light on the deanonymization between real-world heterogeneous social networks. In this paper, we propose a Hybrid De-anonymization Scheme (HDS) aiming at de-anonymizing heterogeneous social networks. HDS firstly leverages the network graph structure to significantly reduce the size of candidate set, then exploits user profile information to identify the correct mapping users with a high confidence. Performance evaluation on real-world social network datasets shows that HDS has considerable accuracy on de-anonymization and significantly outperforms the prior schemes.