11 How Data Mining Can Help Curb City Crime How Data Mining Can Help Curb City Crime Shakir Khan* Abstract : Data mining has been used in different fields to help improve on the efficiency and effectiveness in extracting patterns and predicting certain events using existing statistical data. This paper sets out to apply data mining techniques and approaches in curbing city crime. By obtaining access to crime records held in various city police departments and classifying the data into fields such as address, location, type of crime, and time, it become possible to extract some useful indicators that would help manage crime. Neural networks and other learning algorithms were employed to analyze crime narratives to extract important information that would help the police to identify criminals, crime scenes, times of crime prevalence, and criminal activities in a city. Keywords : Data mining, crime, crime records, Geographic Information Systems (GIS), Curb City Crime, Database Management Systems (DBMS) IJCTA, � International Science Press * Assistant Professor, Information Management Department, College of Computer and Information Sciences Al-Imam Muhammad Ibn Saud Islamic University Riyadh, Saudi Arabia shakirkhan2006@gmail.com 1. INTRODUCTION Conventionally, it has been the entitlement of law enforcement agencies to handle crimes. However, with the proliferation of automated or computerized information systems such as Geographic Information Systems (GIS) and Database Management Systems (DBMS) in cities, governments and private security bodies can now easily trail, unravel and therefore general manage crimes [1][9]. This is because specialized data analysis can aid the law enforcement agencies in this area to speed up the process of identifying, arresting, and prosecuting criminals. To achieve this, though, it requires a collaborative effort between data mining and criminal justice paradigms where the process of clustering models are used to identify crime patterns and if possible, crime hot-spots. Clustering ensures grouping of crimes or hot spots in a particular geographical location such as a city. Through data mining, the clusters help in the identification of the patterns of crime in the city to help deal with common criminals including serial rapists, deadly gangs, and serial killers. Unfortunately, data mining experts have to balance between public and private data that can be extracted and made available for public consumption. This is because of privacy and confidentiality of certain information about crimes, criminals, or victims of crime [2]. In some developed nations such as Singapore, Britain and US, police officers use digital systems to handle crime reporting. Often, the reports have crime-related information grouped into location, crime type, date and time, and crime description among others. While obtaining data based on these categories is possible even through database queries, it is often difficult to extract useful data from crime descriptions because it is stored in form of text. This calls for the application of specialized data mining techniques and models such as Decision Trees and Neural Networks [3][4]. This paper engages in creating a framework for managing crimes in cities using data mining techniques. 9(23), 2016, pp. 483-488