Analysis of criminal networks in Python
Mar 4, 2026·
·
0 min read
Ruslan Klymentiev
Abstract
This webinar introduces criminal network analysis using Python, focusing on co-offending networks derived from real-world police data. Participants will follow the full analytical workflow, from raw police data to the construction and analysis of a social network of co-offenders. The aim is to demonstrate how routinely collected police records can be transformed into network representations that reveal how offenders are connected through shared criminal activity.
The webinar then focuses on individual-level analysis, illustrating how to identify prolific and structurally important offenders. Ego-centric networks are extracted to examine the immediate criminal environment of selected individuals. Centrality measures are introduced and interpreted to assess which actors occupy key structural positions, including potential brokers between groups and individuals who may be particularly relevant for network disruption strategies. Additionally, the webinar introduces the concept of multilayer networks to illustrate how different types of relationships or interactions can be modeled and analyzed.
In addition to network analysis, the webinar highlights practical data manipulation techniques in Python that are commonly required when working with police data. All examples are implemented using open-source Python libraries, including Pandas for data manipulation, NetworkX for network construction and analysis, and PyViz for network visualization. No prior knowledge of social network analysis or Python programming is required, although familiarity with data analysis concepts may be advantageous.
The webinar then focuses on individual-level analysis, illustrating how to identify prolific and structurally important offenders. Ego-centric networks are extracted to examine the immediate criminal environment of selected individuals. Centrality measures are introduced and interpreted to assess which actors occupy key structural positions, including potential brokers between groups and individuals who may be particularly relevant for network disruption strategies. Additionally, the webinar introduces the concept of multilayer networks to illustrate how different types of relationships or interactions can be modeled and analyzed.
In addition to network analysis, the webinar highlights practical data manipulation techniques in Python that are commonly required when working with police data. All examples are implemented using open-source Python libraries, including Pandas for data manipulation, NetworkX for network construction and analysis, and PyViz for network visualization. No prior knowledge of social network analysis or Python programming is required, although familiarity with data analysis concepts may be advantageous.
Date
Mar 4, 2026
Event
International Association of Crime Analysts
Location
Online
