Robbery pattern analysis (RPA) using the concept of multipolarity and examining the influencing factors

Abstract

This paper introduces a method to find out the most robbery-prone time-slot of a place from the past records so as to provide alertness to the public and police. It cannot be predicted who all may be the victims of robbery but can predict the place and time that has high probability of its occurrence. Apart from finding the time-slot where robbery occurrence is more, visualization has been done to draw out the patterns for which race the criminals of robbery belongs. The factors resulting in robbery are also identified and their impact is calculated by developing a regression model so that this information helps to reduce the robbery rate in near future. For this purpose, dataset of Chicago is used as a secondary dataset and to fulfil the objective, the concept of projection is applied on m-Polar fuzzy context with respect to the object set for analysis which will provide the maximum membership-value of the object set depicting the time-slot where more robbery occurrence is there so that more and more security is implemented by the police department.