Data Mining Application in Crime Analysis and Classification

Abstract

The analysis of crime data helps to unravel hidden trends that will aid in better understanding of crime pattern and the nature of those who commits such crimes. It also enables appropriate strategies to be put in place to control such crimes. Literature revealed that data mining has been successfully applied in crime analysis and control. In this work the crime data used was collected from selected Nigerian prisons. Exploratory analysis was performed on the crime data for better insight into the dataset before application of data mining algorithms. The exploratory analyses revealed that majority of the offenders are within the ages of 18 to 34 years old. The mining algorithm was limited to two classification algorithms. C4.5 algorithm was used to classify the data into vulnerable and non vulnerable groups. To verify the reliability of the C4.5 algorithm, Naïve Bayes algorithm was also used to classify the dataset. The result showed that C4.5 classified the data better with higher accuracy of 97% against 93% from Naïve Bayes. The rule generated by the C4.5 classifier revealed that those without educational qualification and not gainfully employed are the vulnerable groups. The authors are of the opinion that education is the way out of crime in Nigeria since most of the offenders either have low educational qualification or none.