Visual Analysis of Implicit Social Networks for Suspicious Behavior Detection

In this paper we show how social networks, implicitly built from communication data, can serve as a basis for suspicious behavior detection from large communications data (landlines and mobile phone calls) provided by communication services providers for criminal investigators following two procedures: lawful interception and data retention. We propose the following contributions: (i) a data model and a set of operators for querying this data in order to extract suspicious behavior and (ii) a user friendly and easy-to-navigate visual representation for communication data with a prototype implementation.