Fuzzy Hypergraph Modeling, Analysis and Prediction of Crimes

Abstract

Hypergraphs are graphs in which more nodes are found in an edge as opposed to two nodes in a simple graph. In this work hypergraphs are created out of crime data and this is used to highlight areas with more crime. Various hypergraph morphological operations like dilation with respect to node, edge, erosion with respect to node, edge are applied which will result in crime data analysis. Moreover, the nodes and edges are fuzzified to make it a fuzzy hypergraph. This is pioneer work which models data using fuzzy hypergraph by applying morphological operations on crime data. Also this is a premier work in multilevel hypergraph. The aim of our work is the development of a novel prediction model which predicts crime behavior of a location using Lukasiewicz implication applied on a fuzzy multilevel hypergraph. Various parameters like proximity to ATMs, highways, shopping malls, railway stations, bus stations, literacy rate, urban/rural factor and the existing crime behavior of a location are considered for this crime prediction.