Towards discovering criminal communities from textual data

In many criminal cases, forensically collected data contain valuable information about a suspect's social networks. An investigator often has to manually extract information from the collected text documents and enter it into a police database for further investigation with criminal network analysis tools. In this paper, we propose a method to discover criminal communities, to analyze the closeness of the members in the communities, and to extract useful information for crime investigation directly from the text documents. The proposed method, together with the implemented software tool, has received positive feedbacks from the digital forensics team of a law enforcement unit in Canada.