Criminal Network Community Detection in Social Media Forensics
Peer reviewed, Journal article
Accepted version
![Thumbnail](/ntnu-xmlui/bitstream/handle/11250/2781464/paper_53.pdf.jpg?sequence=5&isAllowed=y)
View/ Open
Date
2021Metadata
Show full item recordCollections
Original version
10.1007/978-3-030-71711-7_31Abstract
Nowadays, Online Social Networks (OSNs) has created a breeding ground for criminals to engage in cyber–crime activities, and the legal enforcement agencies (LEAs) are facing significant challenges since there is no consistent and generalized framework built specifically to analyse users’ misbehaviour and their social activity on these platforms. Data exchanged over these platforms represent an important source of information, even their characteristics such as unstructured nature, high volumes, velocity, and data inter–connectivity, become an obstacle for LEAs to analyse these data using traditional methods in order to provide it to the legal domain. Although numerous researches have been carried out on digital forensics, little focus has been employed on developing appropriate tools to exhaustively meet all the requirements of crime investigation targeting data integration, information sharing, collection and preservation of digital evidences. To bridge this gap, in our preliminary work we presented a generic digital evidence framework, called CISMO as a semantic tool that is able to support LEAs in detecting and preventing different type of crimes happening on OSNs. This paper gives details of the knowledge extraction layer of the framework. Specially, we mainly focus on analyses criminal social graph structures proving the effectiveness of CISMO in a case study with real criminal dataset. Experimental results reveal that applying appropriate Social Network Analyses (SNA), CISMO framework should be able to query and discover the criminal networks, empowering the criminal investigator to see the connections between people.