Analysis of Crime Data Using Neighbourhood Rough Sets

Analysis of Crime Data Using Neighbourhood Rough Sets

Lydia J. Gnanasigamani (Vellore Institute of Technology, Vellore, India) and Seetha Hari (Vellore Institute of Technology, Vellore, India)
DOI: 10.4018/IJITWE.2020070104
OnDemand PDF Download:
No Current Special Offers


Crime analysis has been carried out to find out patterns and associations in crime incidents. A few of the different latitudes that research has been carried out are the prediction of crime rate, sociological impacts of crime, the contribution of socio-economic factors to the crime and finding the places where the frequency of crime is unusually high. GIS and spatial information have evolved as an inherent part of the crime data as the information is made public by the policing agencies. ‘Crime mapping' refers to mapping a crime to a particular place. Geography or the spatial information of crime plays an important role in the analysis of crime. Previous research have documented the spatial importance in identifying the hotspots and showing crime distribution in a particular geography. This work intends to identify the similarity between regions in the geographical area using Rough Set methodology. By doing so, we can prepare similar crime-fighting strategies for the neighbours and alleviate the crime.
Article Preview

1. Introduction

Criminal incidents or crime has been a constituent of human life since the time they started living together as a family or society. As humans evolved into a more sophisticated life, the types of crimes and the methodology of crime have also grown. The initial interests in the crime were on the incident itself and trying to explain as to why it happened. There has been a lot of research as to why a few people engage in criminal activities while others do not. The advent of psychology and its related sciences have led to multiple theories about individual motivation to crime. These all fall under the bigger heading of criminology, which studies the socio-economic-demographic and psychological background for the crime.

All regions do not experience the same level of crime. Environmental criminology suggests that difference in crime rates in regions may be due to landscape, the access routes and visibility of the area. Routine Activities Theory suggests that for a criminal event to happen the delinquent, the probable victim and the absence of effective control measures must all come together in space and time. Thus space or geography of the event is also of importance. So criminal events need to be studied from multiple dimensions like the law, the offender, the victim and the place.

The very first social scientists, Adolphe Quetelet and A.M. Guerry, were the first to use spatial analysis in the research about crime. Utilising the crime and social data collected from France they were able to determine that across France the criminal activities were crime was not equally distributed and also crime tends to cluster around certain spaces in geography. Many other theorists during that time, in the 1800s, also started using maps to showcase their theories and analysis. In early 1900s single-symbol point maps or pin maps were used by the New York Police Department to exemplify the locations of crime. These pin maps were used to show either all the crimes in a given geographical location or the crimes of a single person or group (serial killers, serial burglars). Around 1920 and 1930s graduated area maps were used in the University of California to show crime and delinquency. All of these maps were hand-drawn and were not accurate.

Large mainframe computers were used to create crime maps around the 1960s and 1970s. Then came the desktop computers with crime mapping facilities. But they were limited in memory and processing speeds. In the 1990s, geographical information systems came into the picture and the desktops speed and memory greatly increased. Today we have a number of GIS supported applications that can be used for crime mapping effectively.

Apart from GIS, a lot of data mining and machine learning algorithms have also been developed and applied to crime data. Initially many of the statistical methods like regression were used to predict the crime rate. Crime pattern analysis to find the different temporal and spatial pattern of crime have also been done for specific localities. Density-based analysis has been done to find the hotspots of crime. Studies have also been carried on about the links between demographic factors, such as population density, immigrant population, socio-economic conditions, income and crime.

Fuzzy sets and Rough sets part of soft computing techniques have also been used to analyse crime data. Fuzzy sets have been used to create offender profiles, analysis of crime data to find anomaly or patterns in crime. Decision support systems for crime data also use rough set approximations. They have also been used for mining association rules. Rough sets have been used for prediction, fraud detection and decision making in crime.

The remainder of the work is organized with section 2 detailing crime mapping and neighbourhood rough sets and its associated literature. The next section discusses the spatial correlation and the use of neighbourhood rough sets as a similarity measure. Followed by the discussion on data, experimental results and discussions and finally the concluding remarks.

Complete Article List

Search this Journal:
Volume 18: 1 Issue (2023)
Volume 17: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 16: 4 Issues (2021)
Volume 15: 4 Issues (2020)
Volume 14: 4 Issues (2019)
Volume 13: 4 Issues (2018)
Volume 12: 4 Issues (2017)
Volume 11: 4 Issues (2016)
Volume 10: 4 Issues (2015)
Volume 9: 4 Issues (2014)
Volume 8: 4 Issues (2013)
Volume 7: 4 Issues (2012)
Volume 6: 4 Issues (2011)
Volume 5: 4 Issues (2010)
Volume 4: 4 Issues (2009)
Volume 3: 4 Issues (2008)
Volume 2: 4 Issues (2007)
Volume 1: 4 Issues (2006)
View Complete Journal Contents Listing