Open data in criminology: a case for synthetic data generation

• In Event: Open Science in Criminology

Abstract

Data sharing in criminology is crucial for collaborations amongst criminologists, cooperation with stakeholder, and the societal impact of criminological research. In addition, data sharing is crucial to enhance trust in research, in that it enables replication of empirical studies that form the basis of policies related to national security issues. However, issues such as privacy concerns, legal restrictions, and reluctance of stakeholders to share sensitive data often hinder comprehensive data sharing. In response to these challenges, we propose the adoption of synthetic data as a viable response to these challenges. Synthetic data generation involves creating artificial datasets that mimic the statistical properties of real data while ensuring individual privacy and confidentiality. Thus, synthetic data can closely resemble real-work dataset without compromising sensitive information. In this presentation, we aim to demonstrate the potential of synthetic data for the field of criminology by discussing its advantages and pitfalls and presenting a proof-of-concept, using a synthetic version of the Dutch Homicide Monitor. Through this presentation, we aim to foster more open and collaborative research in the discipline of criminology.

Authors

• Katharina Krüsselmann, Institute of Security and Global Affairs, Leiden University, the Netherlands

• Marcel Haas, Leiden University Medical Centre

• Marco Spruit, Leiden University Medical Centre & Leiden Institute of Advanced Computer Science

• Marieke Liem, Leiden University