Understanding Domestic Abuse Perpetrators: using unsupervised machine learning to analyse a longitudinal dataset of domestic abuse incidents from Essex Police. A Home-Office funded project. 2022

This project set out to address the following question: Are there any common profiles of domestic abuse perpetrators in Essex, and do they present different risks and opportunities for targeted interventions? To answer this question we conducted a mixed-methods study, analysing a large, longitudinal database of domestic abuse incidents and suspect data -relating to 16,491 suspects and 40,488 observations between 2016-2020- provided by Essex Police. Our findings, presented in this executive summary, reveal at least 4 distinct clusters of domestic abuse suspects in Essex. We explore the implications of these for interventions, training, and commissioning, and make recommendations for further research.