Potential applications of data science in assessing security risk

A/Prof. Garner Clancey1

1University of Sydney, , Australia

This presentation will consider potential applications of data science in understanding and predicting security risk of juvenile detainees. By using state of the art machine learning algorithms and statistical learning, the proposed risk assessment tool is dynamic, individualised and can learn based on large quantities of data. The goal is to estimate an individual’s risk based on longitudinal detainee population data, where risk is defined as the probability of violent incident occurrence within in juvenile justice detention centres.

Better understanding security risks posed by particular detainees potentially improves the operation of juvenile justice centres and makes these safer environments for staff and detainees.

The proposed techniques have not been tested in an operational environment. Consequently, the presentation will highlight potential applications.


Dr Garner Clancey is an Associate Professor in Criminology at the University of Sydney. Before joining the University of Sydney Law School in 2011, Garner worked as a crime prevention consultant (between 2002-2010) and in criminal justice (including Juvenile Justice NSW and the NSW Police Force) and alcohol and other drug agencies in NSW and England (between 1992-2002).