CRIME DATA ANALYSIS WITH ASSOCIATION RULE MINING

Along with the positive developments of the globalizing world, new types of crime such as social media fraud, drug trafficking and vehicle robbery, which have disrupted community welfare and order, have also emerged. With developments in information technology, it is possible to record real-time various data related to subject of crimes, location and time information, type of crime. By analyzing these recorded raw data using various data mining methods, it is possible to extract information that can be used to identify the data or for prediction purposes. In this study, an analysis of the association rules on the NIBRS Crime dataset which includes real crime cases from July 2016 to April 2018 in the state of Maryland in USA was carried out using R program with Apriori algorithm and Rapid Miner with FP-Growth algorithm. With these association rules created, the time intervals, the districts, the types of crimes and the frequency of the occurrences are analyzed and the results of the algorithms are presented. With the results of this analysis; for organizations which are responsible for maintaining the peace and social order, such as security forces and law enforcement agencies; it is possible to follow useful information such as which crimes are committed more frequently and in which time period of day the criminals are more active.