Detecting Pickpocketing Offenders by Analyzing Beijing Metro Subway Data

Nowadays, police are facing with the challenge that crime increases far more rapidly than police force, which generate many problems and cause the police have to seek for support from new technical ways. In this background, the new concept of intelligence-led-policing was proposed. This concept indicated that police should actively identify and analyze the problems by collecting data, such that the potential threatens could be known in early stage and prevention resources would be deployed in more efficient ways. In this paper, a typical work big data analysis in policing was introduced. Pickpocketing offence is a kind of frequently occurred crime in public transportation, and identifying offenders is a great challenge for the police. With nearly a billion of metro subway traffic records accessed from Beijing metro line company, the passengers' routine mobility pattern was analyzed and some pickpocketing suspects were detected by defining some specific rules. The results indicated that significant traveling patterns exist for routine passengers, and some exceptional individuals who has the unique pattern which is quite different from routine passengers are discovered. However, the shortage of the work still exists as the identity of cardholders is anonymous and it cannot be confirmed, but it still provides an example of big data analysis in public security, which might inspire police for further work.