As the rapid development of social media especially in the mobile platform, various homogenous social application has been developed and widely being used(Instagram, Twitter, Facebook, e-mail). Users can share information on the web anytime, anywhere, resulting in large amount of user data. Social network actually have a lot of practical value under the era of big data. Therefore, the data in many fields can be detected by the special instruments such as police departments.

The classic drama PERSON OF INTEREST tells a story that using a backdoor of a “machine” to detect the possibility of the happening of crime, and figure out who might be the potential victim or criminal. The mechanism beneath the “machine” is the analysis of the big data comes from social network of the victim or criminal, for example, if a married man has a mistress, they will try to text each other, the information of their communication will express what are they thinking and planning, so the wife of the married man might become a potential victim.
The big data from personal social network can be used to:
- To reveal hidden relationships, detect any criminal patterns, and prevent security threats.
- Link personal real-time social account to discover unusual user behaviour and suspicious transactions to expose fraud base on the his historical account activities.
- Test new sources of information and various data as evidence of criminal activity, such as the Internet, mobile devices, transaction processing, email, and social media.
In 2010, LAPD (Los Angles Police Department ) become the first employ data technology and information about past crimes to predict future unlawful activity. The technologies they were using are designed to predict where and when the crimes are likely to occur in next 12hours by examining the 10 years of data.
reference:
A new type of database, graph databaes gradually gained it popularity in fighting with digital criminal. Neo4j, one of the commercial graph database company actually assisted the police to find potential criminal groups. Different criminal groups are more likely to know each other (like the small world model). With graph database, it is intuitive to find the SCC of criminals groups and their interaction. Additionally, graph database can find potential victims (like someone suddenly get in touch with detected criminal group) on social media to help police prevent crime from happening.