Social network analysis (SNA) is often used to understand the relationship dynamics. However, researchers have also utilized social network analysis to understand the crime networks in the society. SNA on crime networks can aid in identifying the most influential and central actors within the network. In this blog, I discuss two journal articles that have used SNA for analyzing criminal networks.
In an article “Social Network Analysis of a Criminal Hacker Community”, researchers Lu et al (2010) examine the social organization of a hacker community by analyzing complex and highly structured malicious hacker network called Shadowcrew. Shadowcrew was a hacker group that committed crimes like identity theft and credit card fraud. The study uses text mining and social network analysis methods to investigate the social network structure of this hacker group. The authors state that they utilize SNA for this study as “SNA helps discover the roles and importance of members in a hacker community, potentially providing leverage against harmful activities by a hacker group” (Lu et al, 2010). The research questions are as follows:
1. What is the network centralization of a computer hacker network?
2. Are there members of a hacker network who stand out as critical leaders?
3. How strongly do leaders influence a hacker network?
4. What subgroups exist and interact in a hacker network?
For the data collection, the study uses the data from 182 texts collected from LexisNexis Academia by using the search word “Shadowcrew’. The LexisNexis database included major newspapers, magazines, journals, and law reviews such as The New York Times, The Washington Post, USA Today, Business Week, U.S. Fed News, and Department of Justice Documents. Also, other 25 texts were used in the study which was collected by conducting a search on Google search engine which included open source websites, trial transcripts, and a key court proceedings. By using the text mining tool AutoMap, the study converts the texts into usable meta-matrix data which uses the four major techniques: named-entity recognition, deletion, stemming, and thesaurus creation and application.
Organizational Risk Analyzer and UCINET were used to visualize the network and generate reports. The study includes a total of 23 nodes, 3 of these nodes are isolates. The network centralization looks at the centrality measures at a network-wide level by calculating degree centrality (26.9%), betweenness centrality (4.1%), and closeness centrality (9%) of the network. Considering the low centralization measures, the researchers conclude that Shadowcrew is a decentralized network. Also, to identify the central actors within the network, individual centrality measures including degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality are examined. The results of these suggest that co-founder and administrator of Shadowcrew Andrew Mantovani ranks very high in all the centrality measures, making him an influential and central actor within the network. The researchers also quantify each actor’s connections to any of the thirteen cliques within the hacker network. Analysis of these cliques or sub-networks shows that some hacker members within the network are more directly involved than others. Also, in this decentralized network, the network is organized where each member of the group holds different hierarchical positions with more or less power. These positions are administrators, moderators, reviewers, vendors, and other general members, which suggests that Shadowcrew had an elaborate division of labor.
The second article I came across titled “Social Network Analysis of an Urban Street Gang Using Police Intelligence Data” which aims to systematically understand the gangs and direct law enforcement activities accordingly. The researchers suggest that by utilizing SNA this study can help understand a network of individuals by identifying and mapping their relationships, consequently identifying the key members within the network as well as their associations with other members of the network.
This study was specifically conducted in collaboration with Great Manchester Police in the UK. The study addresses following questions:
1. What can social network analysis tell us about gangs?
2. How useful are the social network analysis outputs for the police?
For the data collection, the police intelligence data of the five individuals who were identified as having gang links (Primary individuals) going back six months were collected. Other individuals (secondary) mentioned in this data were identified. Relationships between primary and secondary individuals were coded by category and input into a database. After repeating the same procedure on secondary individuals, other individuals in their network are identified and the relationships between them are coded according to suspected charged criminal relationship or social links. Where links were criminal, the direction of the relationship was also recorded (whether a crime was committed by person A with person B, committed by A for B, or A had done something to B – e.g. sold drugs to).
They conducted a social network analysis by analyzing the cohesiveness, degree centrality, and betweenness centrality of the actors within the network. The study concludes that by using the SNA on intelligence data, they were able to determine a larger size of the gang-sub network than anticipated as the official reports only contained gang members who were suspected or convicted of a crime. However, by using SNA, they were able to identify individuals who were potentially vulnerable to gang activity. As a result, this would help police departments in deciding where to best use their resources when trying to find gang activity. Researchers suggest that the results of this study have helped Great Manchester Police department to determine gang member links easier than before.
Gunnell, D., Hillier, J., & Blakeborough, L. (2016). Social Network Analysis of an Urban Street Gang Using Police Intelligence Data. Home Office, 89.
Lu, Y., Luo, X., Polgar, M., & Cao, Y. (2010). Social Network Analysis of a Criminal Hacker Community. Journal of Computer Information Systems, 51(2), 31-41.