Development of Risk Assessment Framework for First Time Offenders Using Ensemble Learning

Abstract

Recidivism is generally considered as a deficiency disease in which offenders recommend a crime or repeat an offence. Empirically committing the first crime at a very young age leads to a much higher rebound rate and the continuation of similar offensive behavior. Accordingly, prioritization must be given for the early assessment of recidivism behavior in first-time offender by law enforcing agencies. Different prison studies suggest that recidivism can be curtailed by early behavioral risk assessment in firsttime offenders. Ideally, a psychologist conducts a manual risk assessment using standard psychological assessment tools, which has long been regarded as a standard method for recidivism risk assessment. However, such behavioral examination procedures are usually sluggish and constrained by subjective perceptions. Consequently, this study aims to develop a machine learning-based quantitative risk assessment tool for the recidivism behavioral gradation of first-time offenders. Quantitative gradation and prediction of future recidivism behavior in such offenders are achieved using an ensemble learning model and an advanced machine-learning approach. For the available behavioral data collected from multiple prison locations, simulations were performed, and the experimental results were obtained. It is ascertained that, the proposed three-member and five-member ensemble classifier models lead to 85.47% and 87.72% accuracy respectively in comparison to other standard individual classifiers.

Machine Learning and Criminal Justice: A Systematic Review of Advanced Methodology for Recidivism Risk Prediction

“…Only one study used the Historical, Clinical and Risk Management–20 (HCR-20) with 16 other clinical and non-clinical risk assessment factors in order to determine the likelihood of recidivism among first-time offenders (FTOs) [ 28 ]. Data were collected from various prisons in the Indian state of Jharkhand.…”

“…One study used a generic data standardization and then applied feature selection [ 21 ]. The last one used ANOVA to identify relevant attributes for the current dataset [ 28 ].…”

“…For a better comparison of the studies, it is possible to perform a sorting by the type of recurrence they aim to predict. The sorting leads to four categories: general [ 22 , 23 , 24 , 28 , 31 , 32 ], sexual [ 31 , 32 ], violent [ 22 , 25 , 31 ] and all other recidivism. The last category includes studies that considered a specific type of crime [ 21 ] or referred only to males [ 27 ] or youth [ 26 , 29 , 30 ].…”

“…Table 3 shows the results obtained from the different studies predicting generic recidivism. The ensemble model trained with an HCR-20+ dataset leads to improved performance and seems to be the most effective method in terms of the ACC [ 28 ]. Considering the AUC, the logistic regression yields the highest score with the use of the MnSTARR+ and StatRec datasets [ 22 , 31 ].…”

Criminal behavioral data analysis for recidivation estimation in convicted offenders

“…First time offenders are those who got convicted for any illicit or unlawful act under Indian Penal Code for any crime. This article was based on a study of 220 criminals from the Indian state of Jharkhand [1] who had been convicted at least once in the previous five years. The dataset contains 220 offenders, with a total of 204 participants for the final dataset, as shown in the Table 1 .…”