Towards victim-oriented crime modelling in a social science e-infrastructure.

Area of Science:

• Computational Social Science
• Digital Social Research
• Crime Science

Background:

• The National e-Infrastructure for Social Simulation (NeISS) supports social scientists with advanced computational tools.
• Agent-based modeling is crucial for understanding complex social phenomena like crime.
• Effective crime prevention requires realistic simulations of criminal behavior and victim characteristics.

Purpose of the Study:

• To enhance the agent-based burglary simulation (BurgdSIM) by incorporating individual-level victim data.
• To explore the impact of realistic victim representation on spatial crime patterns.
• To identify demographic and individual factors contributing to burglary risk.

Main Methods:

• Utilizing the UK Digital Social Research programme's e-infrastructure.
• Modifying the existing BurgdSIM agent-based model.
• Integrating individual-level victim data into the simulation parameters.

Main Results:

• The enhanced BurgdSIM model produced significantly different spatial crime patterns compared to the original model.
• Individual-level victim data integration led to a more nuanced understanding of burglary distribution.
• A risk profile analysis identified specific victim characteristics associated with higher burglary risk.

Conclusions:

• Realistic representation of victims is critical for accurate social simulation, particularly in crime modeling.
• The enhanced BurgdSIM model provides a valuable tool for policy-makers in developing targeted crime prevention strategies.
• Further research can leverage this approach to simulate other crime types and social issues.