Criminal identity resolution using social behavior and relationship attributes

We propose a criminal identity resolution technique that utilizes both personal identity and social identity information. Guided by existing identity theories, we examine three types of identity features, namely personal identity attributes, social behavior attributes, and social relationship attributes. We also explore three matching strategies, namely pair-wise comparison, transitive-closure, and collective resolution. Our experiment on synthetic data sets show that both social behavior and relationship attributes improve the performance of identity matching as compared to the use of personal identity attributes alone. The results also show that the collective relational resolution approach outperformed other approaches in terms of F-measure.