Deciphering Dark Patterns: AI and Machine Learning Approaches to Understanding Violent Crime

• In Event: Methodological Issues in the Study of Violence

Abstract

The emergence of 'Big Data', enhanced processing capabilities, and algorithmic advancements has catalyzed the ascendancy of 'machine learning' (ML), a more adaptive and less rigid approach to artificial intelligence (AI) (Hayward & Maas, 2021). ML entails the gradual acquisition of 'correct' or 'useful' rules by the system itself, facilitated by training data rather than explicit programming, as seen in 'expert systems' (Hayward & Maas, 2021). This study employs an AI/ML methodology utilizing extensive databases, with the potential to address various forms of violent crime and other focal points pertinent to criminal justice practitioners, such as human trafficking. The primary objective of this project is to bridge existing gaps in the field by furnishing comprehensive, precise, and centralized data on various forms of violent crime for utilization by students, researchers, and practitioners. Beyond the provision of accessible databases, success metrics will include the regular extraction, evaluation, and validation of fresh data from a select number of crime/news websites utilizing automation and Natural Language Processing (NLP) techniques.

Authors

• Elizabeth Gurian, Norwich University

• Rachel Sickler, Norwich University