Artificial intelligence (AI) has gained attention for its potential to improve law enforcement operations through proactive policing. Advancements in data science have shown the potential benefits of applying machine learning (ML) in the criminal justice sector. Therefore, research in improving methods to forecast the likelihood of criminal reoffending is quickly growing. Creating a cutting-edge model for using ML to predict recidivism is challenging. We picked 12 out of 79 studies from Scopus and PubMed online databases in a comprehensive review that ensures the models can be replicated across various datasets and are suitable for predicting recidivism. Using two specific measures, the 12 research compared different datasets and machine learning algorithms. This study demonstrates that each approach achieves strong performance, with an average accuracy score of 0.81 and an average area-under-the-curve score of 0.74. This systematic research emphasizes essential factors that could enable criminal justice professionals to consistently utilize forecasts of recidivism risk generated by machine learning approaches. The factors include performance indicators, transparent algorithms or explainable AI approaches, and high-quality input data.
Khatun et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: