GIS-Based Multivariate Clustering Analysis of Robbery Crime in Arlington, Texas

Keywords: GIS,Spatial Analysis, Multivariate Clustering, Criminal Justice, Crime Prevention, Violent Crime, Victimization
Abstract Type: Poster Abstract

Abstract

According to the UCR crime summary report provided by the city police department, the number of victims of violent crimes increased by 14.5% between 2018 to 2019 in Arlington, Texas. More than 400 victims were involved in robbery crimes in year 2019. Whether investigating crime or other types of incidents, large volumes of data can make it difficult to identify hidden geographic patterns. This present study attempts to develop a GIS-based multivariate spatial clustering approach to identify how different risk factors and high number of robbery victims are concentrated within computed clusters. This multivariate clustering approach uses a K-means algorithm and machine learning to generate geographic clusters based on similar attributes and proximity. The crime data is provided by the Arlington Police Department for the period of 2017 to 2019. The findings suggest that some clusters are characterized by a concentration of high numbers of restaurants, bars, smoke shops, and recreation facilities with high number of young black male robbery victims. Preliminary results of this study can be helpful to develop tailored strategies to tackle crime at the block group level in suburban areas.