Crime is undesired anti-social behavior and poses serious threat to society. The civilized societies make everything possible to reduce crime within its regime of influence. Alarming the crime prone areas in advance is one of the best strategies for crime to be ceased to happen. The recent socio-economic developments and proliferation of internet technologies have turned the crime into a global phenomenon. In such scenario the crime data to be dealt is huge in volume, diverse in variety and highly location dependent. Hence the contemporary crime data set is highly spatio-temporal in nature where the traditional system of criminal records has failed to maintain the desired level of intelligence and make a substantial prediction. A blend of `Big data' tools for data management and Generalized Linear Regression for statistical analysis is used to draw a useable inference from such time series data set. Such enhancement is supportive to detect similar crime trends among various crime locations for criminal site selection. Consequently ARIMA (Auto Regressive Integrated Moving Average) model affords to minimize the error generated in the predictive model. This research paper aims to locate the offender site in advance with more accuracy. We have explored the Auto Regression Techniques to accurately predict the crime with minimum error for such time series data by identifying the relationship among crimes attributes. The experimental result obtained using 'R' tool show that our formulation work well for all parameters and improves certainty in prediction.
展开▼