As many organizations growth acknowledge the success of data mining, it is being applied to almost every sector to detect problems and provide recommendations to avoid them. Data mining is an overall term for the process architectures to determine and classify the patterns and further to interpret data for the purpose of perceptive actionable trends and preparing plans based on those trends. The data warehousing and data mining applications are being extensively used by organizations for the purpose of increasing sales, business management systems, detecting crime pattern and possibly propose accurate measures to avoid further crime happenings. The useful and effective techniques in data mining tools helps business analysts to extract useful information from large databases collected from years and decades. It is often seem difficult to .Thus data mining is seen as a solution for extracting knowledge from complex or large datasets. Analysts are able to study very large data sets to find all hidden patterns and complex relations by using artificial intelligence, statistical techniques and visualization methods in data mining.
All public safety organisations from many years have been recording useful information of crime taking place in their respective regions. They even attempt to address challenges faced by them associated with astonishing increases in available information, different techniques and better approach to analysis is required.[Richmond Police Department]
Without datamining, effective analysis of information is restricted to small number of aspects that are expected to be important. By digging into the information data mining vastly expands that small analysis scope. While understanding time/place/type-of-crime combinations, organizations can determine high risk of criminal behaviour, in order to undertake more effective preventive measures.
In the context of crime analysis, datamining can be used to model crime detection problems. It can be a challenge for the investigating agencies to deter the increasing amount of crime without using any effective technology like data mining. Agencies have been collecting important data of crime records in different ways. With important information such as date, time, crime weapon, location a techniques in data minining can be used to create data models to predict crime type and frequency in same season and same place.
It is also seen how visualization has become an increasingly important for aspect in today's world. Managements, potential users and analysts are demanding for easy representation of the databases. It becomes easy to discuss when datasets are converted to pictures and animations. Dashboards developed for business purposes dramatically reduce to need for large operational reports and forms. Faster results and decision making are key advantages of effective visualizations. Significant number of different tools are available in the software market to develop representations from datasets. They are based on various technologies, while some tools are proprietary the others are open source. More emphasis is shown on the open source tools as they provide options to change the application according to a developer's perspective.
1.2 Scope and Project Objectives
This will include the detail problems I would intend to tackle. Why the problems are worthy doing/researching and research efforts.
Concentrating on preventing crime in different regions of Washington DC, the following objectives are clearly taken under for research:
To translate the business problem to a data mining problem.
To apply data mining tasks on the dataset and create models with SAS Enterprise Miner.
To recommend measures for particular crime prone areas to decrease the crime rate.
To decrease the high risk type of crime in different regions.
Introduce Law Enforcement rule for the Police Department of Washington DC
To introduce different open source visualizations techniques for any dataset.
Suggest the Police Department for Patrolling forces
1.3 Project Methodology
Methodology is required to design the activities in this data mining project. It will indeed create a set of logical steps to achieve the project objective. For this research the following tools will be used:
SAS Enterprise Miner
SAS Enterprise Guide
Microsoft Excel
Different Open Source tools developed for visualization
1.4 Expected Results
By the end of the research, a recommendation will be provided for the police department of Washington DC to avoid crime in different regions based on the analysis and model evaluations from the dataset.
On the parallel side this research will also include every open source visualization tools released by different vendors and eventually illustrate the flexibility of these tools.
1.5 Chapter Outline
Chapter 1 : Introduction - This Chapter covers the introduction of the topics covered in this research. A background of data mining, crime analysis and crime mapping and its application in crime analysis is discussed. Additionally important points are addressed such as project objectives, project methodology, possible problems that will be tackled .
Chapter 2 : State of the Art - In this section the present researches in crime data mining is covered.
Chapter 3 : Data Minig : An Overview - An overview of data mining and data mining tasks will be discussed. Overall concepts and methods that covers different algorithms and models that can be derived from SAS tools.
Chapter 4 : Crime Data Mining - Over the last few decades the application of data mining is largely used in detecting crime and predicting crime from the databases that investigating agencies are maintaining. This chapter will contain the different levels in crime data mining and how they are used to help solve crime problems. Additionally the presentation part like visualization, crime mapping and GIS Systems are also discussed briefly.
Chapter 5 : Application of Data Mining process in Crime Analysis - Data mining applied on the dataset to evaluate models using appropriate data mining task. This chapter will also cover the business problem of the dataset. A complete walkthrough of data preparation, modelling and data analysis will be shown and explained including the results.
Chapter 6 : Eventually the conclusions of the research will be discussed in this chapter. It will include the findings of the analysis and also recommend for the police department of Washington DC with measures that needs to be taken to avoid or decrease the crime rate in different regions.
Chapter 2 : State-of-the-Art
The researches in Crime Data mining has been significantly increasing after the terror attacks on United States of America in September 2001. Federal agencies like the FBI, CIA and the others are actively maintaining global intelligence to put a stop to future attacks. Since then the data mining technology is believed to be taking big strides to tackle challenges with extraction of information in any datasets.
In 2004, Chen H., Chung W etl., performed a research that would increase efficiency and reducing errors, and that crime data mining techniques can facilitate investigating work and other valuable tasks. In this research they used a Coplink case study to illustrate their crime data mining framework which describes three uses: named-entity extraction, criminal-network analysis and deceptive-identity detection.
In another research, in 2006, the Richmond Police Department also used data mining and crime analysis
Chapter 3 : Data Mining : An Overview
3.1 Concept of Datamining
The Chapter introduces the concept of Data mining, how this technology has evolved in recent years and how successful it has proven for a wide range of industries. The part followed by this will have a deeper insight of a standardized data mining process called CRISP -DM. It will also be discussed how stages in the process are re-used or re-visited for the purpose of the project. Data Mining consists of various different tasks which are applied to potential data sets after understanding the business needs of a project. These tasks are categorised under two techniques of data mining which are called as Predictive Modelling and Descriptive Modelling. Finally the algorithms that are used applied under the tasks are explained in this chapter.
Data Mining (also called data or knowledge discovery) or Knowledge Discovery in Database is the process of analyzing data from different perspectives and summarizing it into logical information - information that can be used to increase revenue, cuts costs, or both. Data mining software is one of a number of analytical tools for analyzing data. It allows users to analyze data from many different dimensions or angles, categorize it, and summarize the relationships identified. Technically, data mining is the process of finding correlations or patterns among dozens of fields in large relational databases.
A very large amount of interest is elicited in the field of data mining, in recent past, not just in research domain but also in commercial domain. The commercial utility of data mining is probably as used or implemented as used in researches. The purposes of using data mining technology covers a wide variety from strategic decision making, wealth generation, analyzing trends to security. The term data mining is also said to be most used to look for hidden patterns and trends in a dataset that is not immediately apparent from summarizing the data.
As you can imagine from the phrase 'Rich Data but Poor Information', the need in today's world is how to extract data from a huge database and transform into logical information. Data Mining has become an answer in these situations where industries want to get a meaningful information on which they can capitalise for their growth. This technology has got increasing acceptance in business and science areas, which needs to work and analyse huge databases and large number of data to discover which they could not have found otherwise.
Data mining uses different algorithms for different types of dataset to serve their respective purpose. After the business problem is identified, the most appropriate algorithm is applied to extract useful information and get meaningful results.
Its roots mainly are based on three important subject areas:
First area is said to be the most vital to make the data mining technology successful which is Statistics. Without it, users can never perform data mining tasks or use any process. Statistics are the foundation of most technologies on which data mining is built. Within the nucleus of today's data mining tools and techniques, statistics plays an important role. Second Area that data mining is also built on is Artificial Intelligence. This has got a different approach than statistics as it applies human-thought-like processing to statistical problems. It applications are mostly found in the high-end commercial product that are used for markets such as government organizations or scientific purposes. This process essentially requires vast computer processing power. The much popular query optimization modules for relational Databases comes from the concept of AI. And finally the third are machine learning, which one can call it a combination of AI and Statistics. The development of all three together combined defines the concept of data mining.
The ultimate goal of Data Mining methods is not to find patterns and relationships as such, but the focus is on extracting knowledge, on making the patterns understandable and usable for decision purposes. Thus, Data Mining is the component in the KDD process that is mainly concerned with extracting patterns, while Knowledge Mining involves evaluating and interpreting these patterns. This requires at least that patterns found with Data Mining techniques can be described in a way that is meaningful to the data base owner. In many instances, this description is not enough, instead a sophisticated model of the data has to be constructed.
Data pre-processing and data cleansing is an vital part in the Data mining process. The process of data mining includes collecting data from different sources at different time points. As they are taken from different places, their integration is not a simple task. Additionally, there may be inconsistency in the data set, erroneous observations and missing values. To assess or measure the quality of the data set is the primary goal for any data mining project or investigation. While an overview of the data is found by simple statistical and table graphics to determine any inconsistencies or errors in the data set some existing features of some data set is also known. It is also worth observing that a big number of organizations still report that most of their task and effort is consumed into supporting the cleaning and transformation of data.
3.2 The evolution of data mining and Current Data Mining Systems:
Timeline for data mining usage over the years.
The evolution of data mining emerged in the 1980's decade and flourished in 1990's. In 1996, Peter Coy from the Business week rightly said that, "He who mines data may strike fool's gold". As industries realised the fruits and advantages of data mining, the usage of this technology did not stop and the organizations continued to collect huge data for data mining purposes.
As this technology was emerging various vendors in the industry developed tools that could do the data mining tasks automatically for the user. The major data mining vendors are seen below
--------------
3.2 Data mining Process
CRISP- DM : CRISP-DM is a standardized process to help run a data mining project. CRISP stand for Cross Industry Standard Process for data mining. Since data mining was a relatively new field everyone was doing their own steps to implement data mining. The pioneers in this discipline together decided to establish one standard process for performing data mining.. CRISP is proved to be appropriate for any industry using any data mining tool. The standard process is essential to make project speedy, well-organized, unfailing and cost effective.
While the sequence of the steps in CRISP-DM is not fixed, the needs of the project may require instant switching from one step to the appropriate random.
CRISP-DM methodology has the following standard process:
Figure 3.a
3.2.1 Business and Data Understanding Process
The first step of CRISP is critical where in the business understanding phase the goals of the project is defined. It is very essential to realize what can we study from the data, what problems would we like the data mining project to solve, what business objectives can be met. In this phase a project plan is created laying out the goals and defining project success. Without clear objective of the project we lack direction and can't gage the project success.
It is very important for the data miner to understand the data well. This phase begins with collecting data or accessing existing data and normally requires experts in this domain. Within understanding of the business need the data are explored with graphically or using basic statistics. What relationships exist in the data. Business and Data understanding are interrelated. Exploring the data and finding relationships can trigger business understanding. Going between the business understanding and the data understanding we form prophesies to test and clear goal for the project.
3.2.2 Data Preparation Process
This phase is usually time consuming, even taking up almost 80% of the project efforts. Raw data is often messy, which means some variable may be missing or incorrect. In this step we must decide how to clean the data.
3.2.3 Estimate and Interpret Model
A wide variety of modelling techniques are available. Additional data preparation may be necessary to properly use a particular algorithm. Therefore the Data Preparation and the modelling processes can go back and forward. The modelling step often generates several predictive data mining models and the models are assessed in this step as well.
3.2.4 Evaluation Process
The models created in the previous steps are reviewed finding the model or collection of models the best achieved the business objectives determined in the business understanding process phase. At the end of this phase, a decision on the use of the data mining results should be reached.
3.2.5 Deployment Process
By this phase we should have a model that meets our business objectives. Deployment uses the model to score new data. This not necessarily is not the end of the project Incorporating new data into the model building phase can improve the performance. Even if the purpose of the model is to increase knowledge of the data, the knowledge gained will need to be organized and presented in a way that the customer can use it.
3.3 Data mining Tasks
Data Mining distinguishes each task in two different types modelling. They are called as Predictive Modelling and Descriptive Modelling. The general tasks used under data mining can be categorized as figure 3.b
Figure 3.b
3.3.1 Descriptive Modelling
This type of modelling can be best imagined as telling a story(http://www.accudata.com/images/pages/WP-Descriptive_v.%20Predictive.pdf)
Descriptive modelling, or clustering, also divides data into groups. With clustering, however, the proper groups are not known in advance; the patterns discovered by analyzing the data are used to determine the groups. For example, an advertiser could analyze a general population in order to classify potential customers into different clusters and then develop separate advertising campaigns targeted to each group. Fraud detection also makes use of clustering to identify groups of individuals with similar purchasing patterns.
3.3.1.1 Cluster Analysis: The term cluster analysis (first used by Tryon, 1939) encompasses a number of different algorithms and methods for grouping objects of similar kind into respective categories. In general, whenever we need to classify a "mountain" of information into manageable meaningful piles, cluster analysis is of great utility.
Cluster Analysis : Cluster analysis is an unsupervised tool which means that no specific target variable is measured. The goal is to simple explore the structure of the data. Cluster analysis sorts the data into similar clusters that show certain charecteristics. In common parlance it is also called look-a-like groups. The simplest mechanism is to partition the samples using measurements that capture similarity or distance between samples.
3.3.1.2 Association Rule
Association rule mining discovers interesting associations and/or correlation relationships among large set of data items. Association rules displays attribute value conditions that occur frequently together in a given dataset. A typical and widely-used example of association rule mining is Market Basket Analysis.
For instance, the data that is collected using bar-code scanners in the supermarkets consists a large number of transaction records. Each record lists all items bought by a customer on a single purchase transaction. Managers would be interested to know if certain groups of items are consistently purchased together. They could use this data for adjusting store layouts (placing items optimally with respect to each other), for cross-selling, for promotions, for catalog design and to identify customer segments based on buying patterns.
Association rules provide information of this type in the form of "if-then" statements. These rules are computed from the data and, unlike the if-then rules of logic, association rules are probabilistic in nature
3.3.1.3 Summarization : At the beginning of each data analysis is the wish and the need to get an overview on the data, to see general trends as well as extreme values rather quickly. It is important to familiarize with the data, to get an idea what the data might be able to tell you, where limitations will be, and which further analyses steps might be suitable. Typically, getting the overview will at the same time point the analyst towards particular features, data quality problems, and additional required background information. Summary tables, simple univariate descriptive statistics, and simple graphics are extremely valuable tools to achieve this task.
3.3.1.4 Anamoly Detection : The set of data points that are considerably different than the remainder of the data. This method is said to be unsupervised where validation can be challenging just like for clustering. The general steps in Anomaly detection includes building a profile of the normal behaviour, where profile can be patterns or summary statistics for the overall population. Further step includes use of normal profile to detect anomalies where anomalies are observations whose characteristics differ significantly from the normal profile. (http://www-stat.stanford.edu/~jtaylo/courses/stats202/notes/chap10_anomaly_detection.pdf) The types of anomaly detection are as follows: Graphical & Statistical based, distance based and model based.
Fig : 3.3.1.4.a
3.3.2 Predictive Modelling
Predictive Modelling is usually applied to identify data mining projects with the goal to identify a statistical or neural network model or set of models that can be used to predict some response of interest. For example, a credit card company may want to engage in predictive data mining, to derive a (trained) model or set of models (e.g., neural networks, meta-learner) that can quickly identify transactions which have a high probability of being fraudulent.
Predictive models can be used for filtering, analysis and prediction
3.3.2.1 Classification
Classification is a data mining (machine learning) technique used to predict group membership for data instances. For example, you may wish to use classification to predict whether the weather on a particular day will be "sunny", "rainy" or "cloudy". Popular classification techniques include decision trees and neural networks.
3.3.2.2 Regression : The regression functions are used to determine the relationship between the dependent variable (target field) and one or more independent variables. The dependent variable is the one whose values you want to predict, whereas the independent variables are the variables that you base your prediction on.
A Regression Model defines three types of regression models: linear, polynomial, and logistic regression. The model Type attribute indicates the type of regression used. Linear and stepwise-polynomial regression are designed for numeric dependent variables having a continuous spectrum of values. These models should contain exactly one regression table. The attributes normalization Method and target Category are not used in that case. Logistic regression is designed for categorical dependent variables. These models should contain exactly one regression table for each target Category. The normalization Method describes whether/how the prediction is converted into a probability.For linear and stepwise regression, the regression formula is:
Dependent variable = intercept + Sumi (coefficienti * independent variablei ) + error
For logistic regression the formula is:
y = intercept + Sumi (coefficienti * independent variablei )
p = 1/(1+exp(-y))
p is the predicted value.
Regression is a data mining function that predicts a number. Age, weight, distance, temperature, income, or sales could all be predicted
using regression techniques. For example, a regression model could be used to predict children's height, given their age, weight, and other
factors.
A regression task begins with a data set in which the target values are known. For example, a regression model that predicts children's height could be developed based on observed data for many children over a period of time. The data might track age, height, weight, developmental milestones, family history, and so on. Height would be the target, the other attributes would be the predictors, and the data for each child would constitute a case. In the model build (training) process, a regression algorithm estimates the value of the target as a function of the predictors for each case in the build data. These relationships between predictors and target are summarized in a model, which can then be applied to a different data set in which the target values are unknown.
Linear Regression
The simplest form of regression to imagine is linear regression with a single predictor. The technique of linear regression can be used if the relationship between x and y can be approximated with a straight line, as shown in Figure 4-1.
Image of Liner relationship between X and Y
Often the relationship between x and y cannot be approximated with a straight line. In this case, a nonlinear regression technique may be used. Alternatively, the data could be preprocessed to make the relationship linear.
Image of Non-Liner relationship between X and Y
3.3.2.3 Prediction :
3.3.2.4 Time Series Analysis
3.3.1 Algorithms
Chapter 4 : Crime Data Mining
In this chapter the application of data mining is shown. Following part explains more about Crime Analysis, the different levels in crime analysis and how it is useful with data mining for helping prevent crime levels. The volume of crime data is increasing along with the incidence and complexity of crimes.
Data mining tools are powerful tools that criminal investigators who may lack extensive training as data analysts can use to explore large databases quickly and efficiently.
Crime investigators must apply a spectrum of techniques to discover associations, identify patterns and make predictions.
4.1 Crime Analysis
The term crime analysis is defined as a set of analytical and systematic processes organized for providing up-to-date and appropriate information corresponding to crime patterns and trend correlations to facilitate the administrative and operational individual in planning the deployment of resources for the prevention and suppression of criminal activities, aiding the investigative process, and increasing apprehensions and the clearance of cases.
It is indeed true that every investigating agency should have crime analysis capability. Or on the other hand it should maintain such data like the one used in this research to make it available to the analysts in data mining to help prevent, reduce, and solve crime and other crime possibilities.
Within this context, Crime Analysis supports a number of department functions including patrol deployment, special operations, and tactical units, investigations, planning and research, crime prevention, and administrative services (budgeting and program planning). --Steven Gottlieb et al., 1994, "Crime Analysis: From First Report To Final Arrest."
Advances in technology, which allow analyses of large quantities of data, are the foundation for the relatively new field known as crime analysis. Crime analysis is an emerging field in law enforcement without standard definitions. This makes it difficult to determine the crime analysis focus for agencies that are new to the field. In some police departments, what is called "crime analysis" consists of mapping crimes for command staff and producing crime statistics. In other agencies, crime analysis might mean focusing on analyzing various police reports and suspect information to help investigators in major crime units identify serial robbers and sex offenders. Some analysts do all this and other types of analysis. The role of the crime analyst varies from agency to agency.
Crime analysis is the act of analyzing crime. More specifically, crime analysis is the breaking up of acts committed in violation of laws into their parts to find out their nature and reporting statements of these findings. The objective of most crime analysis is to find meaningful information in vast amounts of data and disseminate this information to officers and investigators in the field to assist in their efforts to apprehend criminals and suppress criminal activity. Assessing crime through analysis also helps in crime prevention efforts. Preventing crime costs less than trying to apprehend criminals after crimes occur.
There are different levels that crime analysis is measured:
4.1.1 Tactical Level
Tactical crime analysis involves with immediate criminal offenses to promote quick response. It provides information to assist operational personnel in the identification of specific crime trends and in the arrest of criminal offenders. The primary goal is to identify crime trends and patterns/series. This type of crime analysis level focuses on and will work closely with patrol officers and investigators. Suggestions for tactical responses, such as increasing patrol in a certain area at a specific time period,may be part of a crime analysis report or bulletin. Tactical crime analysis responsesmay also include altering the environment to prevent crimes. The concept of crime prevention through environmental design (CPTED) stresses changing the environment to prevent crime after careful analysis of environmental factors contributing to crime problems. A few examples of environmental factors contributing to burglaries might include poor lighting and overgrown shrubbery, accessibility for criminals to major thoroughfares as escape routes, and ineffective door locks. Some analysts, concerned about specific chronic or emerging crime problems, visit the physical areawhere the crimes have occurred to assess possible environmental contributing factors
4.1.2 Strategic Level
This level is more into operational strategies and which looks for solutions to existing-continuous problems. It provides information for resource allocation purposes, including patrol scheduling and beat configuration. Its purpose is to identify unusual crime activities over certain levels or at different seasonal times, identify unusual community conditions, provide police service more effectively and efficiently by matching demands for service with service delivery, reduce and/or eliminate recurring problems, and assist in community policing or problem-oriented policing. Strategic crime analysis is concerned with long-range problems and planning for long-term projects. Strategic analysts examine long-term increases or decreases in crime, known as "crime trends." A crime trend is the direction of movement of crime and reflects either no change or increases/decreases in crime frequencies within a specific jurisdiction or area. For example, strategic analysts might study increased car thefts during the winter months when citizens warm up their cars, leaving them unlocked and unattended in various locations. Another example would be strategizing a plan of attack for decreasing "open-garage-door burglaries," which often take place during the summer when citizens leave their garage doors standing open for long periods of time, inviting burglars to walk into the structure and help themselves to expensive items. Strategic crime analysts may provide information to crime prevention officers, community-oriented policing officers, planning and research, and community outreach programs. Together, the groups can work to identify overall increases in specific areas of crime, develop an action plan to address each issue, and work together to decrease the overall crime in particular areas, developments, complexes, business districts, or in the jurisdiction in genera
4.1.3 Operational/ Administrative Level
involves long range projects. Tasks include providing economic, geographic and law enforcement information to police management, city hall, city council, and neighborhood/citizen groups. Its purpose is financial, organizational, political, and legislative. It is critical to budget, personnel, public information, and legal issues. Administrative crime analysis focuses on providing summary data, statistics, and general trend information to police managers. This type of analysis involves providing descriptive information about crime to department administrators, command staff, and officers, as well as to other city government personnel and the public Administrative crime analysis provides a range of services to a range of customers. Administrative crime analysis information can be significantly automated using technological resources. The automation of traditional administrative crime analysis tasks allows crime analysts to focus more on using their time and skills for tactical and strategic crime analysis
4.2 Crime Mapping, GIS Systems & Visualizations
Crime mapping is used by analysts in law enforcement agencies to map, visualize, and analyze crime incident patterns. It is a key component of crime analysis and the CompStat policing strategy. Mapping crime, using Geographic Information Systems (GIS), allows crime analysts to identify crime hot spots, along with other trends and patterns.
Using GIS, crime analysts can overlay other datasets such as census demographics, locations of pawn shops, schools, etc., to better understand the underlying causes of crime and help law enforcement administrators to devise strategies to deal with the problem. GIS is also useful for law enforcement operations, such as allocating police officers and dispatching to emergencies.
4.3 Law Enforcement
Law enforcement is the combined name for professionals who are dedicated to upholding and enforcing the laws and statutes that are currently in force in a given jurisdiction. There are law enforcement jobs that focus on local settings, while others are focused more on upholding and enforcing national laws. In addition to enforcing laws, the function of legal enforcement also involves managing the punishment process for people who are convicted of crimes, up to and including managing the process of incarceration.
References
MC Cue C, Data Mining and Predictive Analysis: Intelligence Gathering and Crime Analysis Colleen, Butterworth-Heinemann; 1 edition (May 1, 2007)
Matignon R., Data Mining Using SAS Enterprise Mine, Wiley-Interscience (August 3, 2007)
Cios K. J., Pedrycz W., Data Mining: A Knowledge Discovery Approach, Springer; 1 edition (February 1, 2007)
Nath S V, Crime Pattern Detection Using Data Mining, Oracle Corporation,
ESRI, Crime Analysis: GIS Solutions for Intelligence-Led Policing, 2008
Chen H, Chung W, Jie J, XuCrime Data Mining:A General Framework and Some Examples, IEEE Computer Society, 2004
McCue C, Data Mining and Predictive Analytics: Battlespace Awareness for the War on Terrorism, Defense Intelligence Journal, 13-1&2 (2005)
McCue C, Data Mining and Crime Analysis in the Richmond Police Department, SPSS, 2006
Johnson C.P., Crime Mapping and Analysis Using GIS, Geomatics 2000: Conference on Geomatics in Electronic Governance, Pune, January 2000
Gottlieb S., et. al. Crime Analysis: from First Report to Final Arrest, Alpha Publishing, 1994
Harries, K. Mapping Crime: Principle and Practice, CMRC, 1999
Crime Analysis and Crime Mapping Information Clearinghouse
Goldsmith, V.Analyzing Crime Patterns: Frontiers of Practice. Sage: Thousand Oaks, CA, 2000
Reuland, M., Information Management and Crime Analysis: Practitioners' Recipes for Success. Police Executive Research Forum: Washington, DC, 1997
Springer, S.S. Crime Prevention and Control: Index and Analysis of Activity, Success and Services. ABBE: Washington, DC, 1992
COMPASS, Community Mapping, Planning, and Analysis for Safety Strategies,Program Brief. National Institute of Justice: Washington, DC, 2000
Harries, K., Geographic Factors in Policing. Police Executive Research Forum: Washington, DC, 1990
Harries, K., Mapping Crime: Principle and Practice. US Department of Justice, Office of Justice Programs: Washington, DC, 1999
LaVigne, N., Wartell, J., Crime Mapping Case Studies: Successes in the Field, Volume 2. Police Executive Research Forum: Washington, DC, 2000
Dutta, M., Hussain, Z. Determinants of crime rates: Crime Deterrence and Growth in post-liberalized India, Munich Personal RePEc Archive, 2009
Weisburd, D., McEwen, T., Crime Mapping and Crime Prevention, Criminal Justice Press Monsey, New York, U.S.A. 1997
Shang, W., Dong, H., Intelligent Decision-making System Based on Data Mining, International Joint Conference on Computational Sciences and Optimization, 2009
Luo, J., Chen, D., Mapping Rules Based Data Mining for Effective Decision Support Application, International Seminar on Business and Information Management IEEE, 2008
Brown, D. E., Hagen, S., Data association methods with applications to law enforcement, Decission Support Systems, Elsevier Science B.V., 2002
Redmond, M.,Baveja, A., A data-driven software tool for enabling cooperative information sharing among police departments, European Journal of Operational Research, Elsevier Science B.V., 2002
Vladimir Estivill-Castro and Ickjai Lee, Data Mining Techniques for Autonomous Exploration of Large Volumes of Geo-referenced Crime Data
Ester, M., Kriegel, H. -P., Sander, J. and Xu, X., "A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise", Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining, pp. 226-231., 1996
Brown, D. E.,Hagen. S.C., Correlation Analysis for Decision Support with Applications to Law Enforcement, IEEE, 1999
Chen H., Wang F., Artificial Intelligence for Homeland Security,IEEE Computer Society, 2005
