Lizzy – Building an AI-powered domestic abuse risk assessment tool based on nationally representative online survey data

This research aims to develop a novel domestic abuse risk assessment tool, called Lizzy, for predicting the victimisation of German female victims of physical intimate partner violence (IPV). Our approach includes actuarial and machine learning techniques based on data from a longitudinal online anonymous survey with a nationally representative sample of 3,878 respondents (July to November 2023). Four algorithms were employed: CatBoost, XGBoost, Logistic Regression, and Random Forest. Logistic regression performed best with an accuracy of 0.80 and AUC of 0.85. We find that both non-physical and physical predictors contribute to the performance of the models.