1Yong Seog Kim, Professor, Data Analytics and Information Systems Department, Utah State University, United States of America.
2Erin Crump, Department Head, Data Analytics Department, Bridgerland Technical College, United States of America.
*Corresponding Author Yong Seog Kim, Professor, Data Analytics and Information Systems Department, Utah State University, United States of America. Email: yong.kim@usu.edu
The worldwide prevalence of drug overdose and the misconception on psychotropic substances lead to the increased incidents of drug use disorders, drug offences and environmental harms along with financial burden on local and federal government for drug control and prevention. As a small step to reduce drug-related offences, we analyze the data sets consisting of drug- or alcohol-related crime incidents to discover temporal and seasonal patterns of such crimes. More importantly, we employ a density-based clustering algorithm to find a natural grouping of the geographic locations of crime incidents based on their longitude and latitude information. By visualizing such clusters with major crime types for each cluster, we allow residents and public safety officers to easily identify hot spots of drug-related crimes and hence develop new prevention plans to cope with drug-related crimes.
Clustering, DBSCAN, Drug Crimes, Geospatial Analysis, Psychotropic Substances
