Techniques for the identification of semantically-equivalent online identities

Abstract

The average person today is required to create and separately manage multiple online identities in heterogeneous online accounts. Their integration would enable a single entry point for the management of a person's digital personal information. Thus, we target the extraction, retrieval and integration of these identities, using a comprehensive ontology framework serving as a standard format. A major challenge to achieve this integration is the discovery of semantic equivalence between multiple online identities (through attributes, relationships, shared posts, etc.). In this paper we outline a hybrid syntactic/semantic-based approach to online identity reconciliation. We also discuss the results of syntactic matching experiments conducted on real data, the current status of the work and our future research and development plans in this direction. © Springer-Verlag 2013.