A Review on a Classification Framework for Supporting Decision Making in Crime Prevention

Abstract

Increasing volume of crimes has brought a serious problem to many countries across the world. Crime prevention is an important component of an overall strategy to reduce crime and to strengthen public safety. Although, Supporting Decision Making (SDM) in crime prevention is an important topic but a comprehensive literature review on the subject has yet to be implemented. Thus, this study presents a systematic and comprehensive review on a classification framework for SDM in crime prevention. Forty four journal articles on the subject published between 2000 and October, 2015 were analyzed and classified into two categories of index crime (violent crime and property crime) and six classes of data mining techniques (prediction, classification, visualization, regression, clustering and outlier detection). The results of this study clearly show that data mining especially prediction and clustering techniques have been applied most extensively in both index crime categories. The main data mining techniques used for SDM in crime prevention are Bayesian, neural network and nearest neighbor. This study also addresses the gaps between SDM in crime prevention and the needs of practitioners to encourage more researches in crime analysis. Finally, it concludes with some suggestions for future research on SDM in crime prevention.

Themes in data mining, big data, and crime analytics

“…The latter two points there has not been much development. (Mohamad Noor et al, 2015) review of 44 crime articles (2000–2015) classified into six classes of data mining techniques (prediction, classification, visualization, regression, and clustering and outlier detection) noting that prediction and clustering techniques were the most prolific, with the top technologies being Bayesian, neural network, and nearest neighbor.…”

A Systematic Literature Review of Fraud Detection Metrics in Business Processes

“…Data mining-based methods process large quantities of data to derive an underlying meaning [42]. These methods can be categorized into six main categories: classification, clustering, regression, outlier detection, visualization, and prediction [43], [44], as cited in [12]. In addition, data mining methods combine multiple techniques from other domains such as statistics, machine learning, high-performance computing, and many application domains [45].…”