Book Title: Computational Criminology: AI Applications in Forensic Science and Criminal Justice
Editors: Dr. Xavier Louis, Dr. Surbhi Girdhar, Ms. Aswathi Chandran Nair, Mr. Ravi Kumar, and Ms. Nandini Katare
Chapter: 7
DOI: https://doi.org/10.59646/704/7
Author: Sree Durga Madhu
Abstract
Crime is not merely spatial; it is also profoundly temporal. Offences cluster in time, recidivism unfolds across measurable trajectories and the rhythms of urban life work, school, leisure, sleep shape both the opportunity structures of offending and the windows in which preventive intervention is possible.[1] For most of the twentieth century, however, criminological research treated time as a covariate to be controlled rather than a structure to be modelled. Cross-sectional regressions estimated the average effect of a predictor on a binary outcome, leaving the rich temporal architecture of criminal behaviour largely opaque. The computational turn has begun to restore time to its proper place at the centre of criminological analysis.
Sequence-aware machine learning drawn from recurrent neural networks, temporal convolutional architectures, attention-based transformers, and survival-analytic neural models has reshaped the practice of temporal crime forecasting over the past decade. These methods, originally developed for natural language, speech recognition, and physiological signal processing, are now routinely applied to streams of criminal events, police-citizen contacts, court appearances, and supervision contacts. They do not merely predict whether an event will occur; they predict when, in what order, and conditional on what preceding history. They offer, in principle, a more honest representation of criminal careers than the static actuarial scales that have dominated correctional practice since the 1920s.[2]
[1]Andrew V Papachristos, “The Coming of a Networked Criminology” in John MacDonald (ed), Measuring Crime and Criminality (Transaction, 2011) 101, 103.
[2]Lawrence E Cohen and Marcus Felson, “Social Change and Crime Rate Trends: A Routine Activity Approach” (1979) 44 American Sociological Review 588, 590.