Cluster based Zoning of Crime Info

ABSTRACT

The criminal behavior is a disorderliness that is a combined result of social and economic aspects. The crime rate has expanded and the activities of criminals have broaden in last few decades due to better communication system and transport. Crimes cause terror and damage our community enormously in several means. In cities and towns the crime trends rises due to fast developmental activities and increase in population. In India, the regional location has a powerful impact on criminal activity. The CrimeInfo report of National Crime Records Bureau (NCRB), India collects, analyze and publish the crime data. The crime profiling and zoning can be modeled with utilization of data mining. In this paper, we make cluster analysis by using k-means cluster algorithm on criminal dataset of India. The cluster input is used to create custom India map with the cluster zones of states. The custom maps displays an overall crime profiles of states which helps police and law enforcement department to take additional preventive measures to combat against the crime and plan advanced investigation strategies. The crime trend and zoning knowledge can also be helpful in cautioning police to increments and reductions in levels of actions.

INTRODUCTION

In current era, criminals have maximal utilization of all modernized innovations and novel practices in perpetrating the crimes. Worldwide top priority is given by all government departments towards security and curtails the crime occurrence. A crime trend is a continuous, long-term rise or fall in temporally-based information. The crime records assumes a critical part in the planning of police working for control and discovery of crime. The Indian Police, throughout the years, have tried to enhance the proficiency of the crime records systems to perform their duties with higher productivity and adequacy. The automation of criminal records and the Police Computer Network system has delivers huge criminal information. The National Crime Records Bureau NCRB [1], home affairs ministry, India collects & maintains criminal data and publish reports of crime statistics documents. The crime data could be analyzed to interpret the emerging crime trends at high quality both locally and nationally.

In this paper we use cluster technique of k-means algorithm on criminal dataset of India for crime analysis The crime dataset is developed by applying complex query on the CrimeInfo India database [1]. The crime dataset is inputted into WEKA software to construct cluster zones based on k means clustering method. The cluster technique builds a model of states with high, medium and low crime zones. The cluster output of WEKA is passed manually as input to MyCustom map [2], an online interactive map tool of maps of India to create custom India map with the cluster zones of states. The custom maps displays an overall crime profiles of states which helps police and law enforcement department to take additional preventive measures to combat against crime and plan advanced investigation strategies. The crime trend and zoning knowledge can also be helpful in cautioning police to increments and reductions in levels of preventive actions.

METHODOLGY

 The CrimeInfo ver1.0 is a database software [1] developed by NCRB, India with the cooperation of the UN system. The crime dataset is developed by applying complex query on the CrimeInfo, India database.

The dataset obtained from NCRB does not contain any class label. To make groups of sets of the raw data it is better to use cluster method. Clusterization is a concept of data mining which groups objects into a category based on some attributes. WEKA software is a tool to analysis the historical data with a collection of machine learning and data mining algorithms [16].

The crime dataset is inputted into WEKA software to construct cluster zones using K-means cluster method. The Kmeans clustering [16] algorithm groups objects depending on characteristics. The groups are developed by minimizing sum of squares of distances between data and concerned group centroid. Fig 1 shows k-means clustering algorithm.

The cluster output of WEKA is passed manually as input to MyCustom map [2], an online interactive map tool of maps of India to create custom India map with the cluster zones of states. The methodology adopted for Crime Zoning of Crime Info is shown in Fig 2 Fig. 1. K-means clustering algori

CONCLUSION AND FUTURE WORK

Criminology is a sensitive area where proficient clustering approaches of data mining plays vital role for crime analysts. We focus on criminal analysis by executing k-means clustering algorithm on CrimeInfo NCRB dataset of India. The clustering method is used to group the states in India according to the criminal data of total, male and female number of crimes for the year 2010.

The cluster zoning and custom maps generated can help state police and law enforcement department to take additional preventive measures in high and medium crime risk zones to combat against crime and plan advanced investigation strategies. The crime trend and zoning knowledge can also be helpful in cautioning police to increments and reductions in levels of preventive actions.

 In future, we can likewise perform different methods of data mining on CrimeInfo NCRB India dataset with more criminal attributes to identify crime trend, crime patterns. The knowledge can be used to frame crime control methods and optimal deployment of resource in crime prevention for future.