Survey of Data Mining Techniques on Crime Data Analysis

Integrated Intelligent Research (IIR) International Journal of Data Mining Techniques and Applications
Volume: 01 Issue: 02 December 2012 Page No.47-49
ISSN: 2278-2419
47
Survey of Data Mining Techniques on Crime Data
Analysis
Abstract -Data mining is a process of extracting knowledge
from huge amount of data stored in databases, data warehouses
and data repositories. Crime is an interesting application where
data mining plays an important role in terms of prediction and
analysis. Clustering is the process of combining data objects
into groups. The data objects within the group are very similar
and very dissimilar as well when compared to objects of other
groups. This paper presents detailed study on clustering
techniques and its role on crime applications. This study also
helps crime branch for better prediction and classification of
crimes.
Key words: Data mining, Crime data analysis, clustering.
I. INTRODUCTION
In recent years, volume of crime is becoming serious problems
in many countries. In today’s world, criminals have maximum
use of all modern technologies and hi-tech methods in
committing crimes [1]. The law enforcers have to effectively
meet out challenges of crime control and maintenance of
public order. Hence, creation of data base for crimes and
criminals is needed. Developing a good crime analysis tool to
identify crime patterns quickly and efficiently for future crime
pattern detection is challenging field for researchers.Data
mining techniques have higher influence in the fields such as,
Law and Enforcement for crime problems, crime data analysis,
criminal career analysis, bank frauds and other critical
problems.
In recent years, data clustering techniques have faced several
new challenges including simultaneous feature subset
selection, large scale data clustering and semi-supervised
clustering. Mostly, cluster analysis is an important human
activity which indulge from childhood when learn to
distinguish between animals and plants, etc by continuously
improving subconscious clustering schemes. It is widely used
in numerous applications including pattern recognition, data
analysis, image processing, and market research etc [16].
Recent researches on these techniques link the gap between
clustering theory and practice of using clustering methods on
crime applications [17]. Cluster accuracy can be improved to
capture the local correlation structure by associating each
cluster with the combination of the dimensions as independent
weighting vector and subspace span which is embedded on it
[14,15].
The organization of the paper is as follows. Section II
describes methods used in crime domain. Role of
preprocessing in data mining is presented in section III.
Clustering methods have discussed in section IV and section V
concludes the paper.
II. DATA MINING ON CRIME DOMAIN
Recent developments in crime control applications aim at
adopting data mining techniques to aid the process of crime
investigation. COPLINK is one of the earlier projects which is
collaborated with Arizona University and the police
department to extract entities from police narrative records [9].
Bruin, Cocx and Koster et al. presented a tool for changing in
offender behavior. Extracted factors including frequency,
seriousness, duration and nature have been used to compare the
similarity between pairs of criminals by a new distance
measure and cluster the data accordingly [2].
Brown proposed a framework for regional crime analysis
program (ReCAP) [1]. The data mining was adopted as an
algorithm for crime data analysis. J.S.de Bruin et.al compared
all individuals based on their profiles to analyze and identify
criminals and criminal behaviors [2]. Nath et.al used K-means
clustering to detect crime pattern to speed up the process of
solving crimes [5].
Adderley and Musgrove applied Self Organizing Map (SOM)
to link the offenders of serious sexual attacks [11]. Recently,
Ozgul et.al proposed a novel prediction model CPM (Crime
Prediction Model) to predict perpetuators of unsolved terrorist
events on attributes of crime information that are location, date
and modus operandi attributes [7]. LianhangMa, Yefang Chen,
and Hao Huang et.al presented a two-phase clustering
algorithm called AK-modes to automatically find similar case
subsets from large datasets [13]. In the attribute-weighing
phase, the weight of each attribute related to an offender’s
behavior trait using the concept Information Gain Ratio (IGR)
in classification domain is computed. The result of attribute-
weighing phase is utilized in the clustering process to find
similar case subsets.
III. ROLE OF DATA PREPROCESSING
Data preprocessing techniques are mainly used for producing
high-quality mining results. Raw data are being preprocessed
before mining because data are in different format, collected
from various sources and stored in the data bases and data
warehouses. Major steps involved in data mining are data
Revatthy Krishnamurthy, J. Satheesh Kumar
Research Scholar in Computer Science,Bharathiar University, Coimbatore
Asst. Professor in Computer Applications, Bharathiar University, Coimbatore
Email: revatthykm11@yahoo.com, jsathee@rediffmail.com

ISSN: 2278-2419
48
cleaning, data integration, data transformation and data
reduction have shown in Figure 1.
Figure 1. Data preprocessing
In the data cleaning step, missing values will be filled, noisy
data will be smoothened, outlier’s data will be removed and
inconsistent data are resolved. Data integration step undergoes
merging of data. Data normalization and attribute construction
are done in the data transformation for standardizing data. The
data range falls under 0.0 to 1.0 during standardization.
Attribute subsets are selected from large dataset and
dimensionality has reduced. After preprocessing, finally
standard data underwent the process of mining and hence
better results have obtained.
IV. CLUSTERING METHODS FOR CRIME
DOMAIN
The partition clustering methods primarily classified into K-
means, AK-mode and Expectation-Maximization algorithms.
The partitioning method constructs ‘k’ partitions of the data
from a given dataset of ‘n’ objects. After preprocessing, the
operational data are undergoing the clustering techniques for
grouping objects as different clusters.
A. K-means Clustering Algorithm
K-means algorithm mainly used to partition the clusters based
on their means. Initially number of objects are grouped and
specified as k clusters. The mean value is calculated as the
mean distance between the objects. The relocation iterative
technique which is used to improve the partitions by moving
objects from one group to other. Then number of iterations is
done until the convergence occurs. K-means algorithm steps
are given as
Input: Number of clusters.
Step1: Arbitrarily choose k objects from a dataset D of N
objects as the initial cluster centers.
Step 2: reassign each object which distributed to a cluster
based on a cluster center which it is the most similar
or the nearer.
Step 3: Update the cluster means, i.e. calculate the mean
value of the object for each cluster.
Output: A set of k clusters.
K-means algorithm is a base for all other clustering algorithms
to find the mean values.
B. Ak-mode Algorithm
Ak- mode clustering algorithm is a two step process such as
attribute weighing phase and clustering phase. In the attribute
weighing phase, weights of the attributes are computed using
Information Gain Ratio (IGR) value for each attribute. The
greatest value of weight is taken as decisive attribute. The
distance between two categorical attributes is computed as the
difference between two data records gives the similarity
measures. The analyst has set the threshold value α with the
help of the computation result of similarity measures. This
algorithm is mainly used for categorical attributes. Ak-mode
algorithm steps are as follows:
Input: Data set, weighted attributes and threshold value.
Output: cluster result
Step1: Find the number of clusters k and find initial mode of
every cluster.
Step2: Calculate the distance for every mode and its
closest mode.
Step3: update each cluster mode.
Step4: this process terminates when all the modes do not
change. Else go to step 2.
Ak-mode algorithm has been used to find the similar subsets
automatically from large datasets and mainly applied for
categorical attributes.
C. Expectation-Maximisation algorithm
Expectation- Maximization algorithm is an extension of K-
means algorithm which can be used to find the parameter
estimates for each cluster. The entire data is a mixture of
parametric probabilistic distribution. The weight of attributes is
measured in the probability distribution and each object is to be
clustered based on the weights instead of assign the objects to
the dedicated clusters in K-means. To find parameter estimates,
the two steps of iterative refinement algorithm are used.
Step1: Expectation step:
For each object of clusters, this step calculates the probability
of cluster membership of object xi.
Step2: Maximization step:

ISSN: 2278-2419
49
Re-estimate or refine the model parameters using probability
estimation from step1.
This EM algorithm is easy to implement and it converges fast
in practice.
V. CONCLUSION
Crime data is a sensitive domain where efficient clustering
techniques play vital role for crime analysts and law-enforcers
to precede the case in the investigation and help solving
unsolved crimes faster. Similarity measures are an important
factor which helps to find unsolved crimes in crime pattern.
Partition clustering algorithm is one of the best method for
finding similarity measures. This paper deals detailed study
about importance of clustering and similarity measures in
crime domain.
REFERENCES
[1] D.E. Brown, "The regional crime analysis program (RECAP): A Frame
work for mining data to catch criminals," in Proceedings of the IEEE
International Conference on Systems, Man, and Cybernetics, Vol. 3, pp. 2848-
2853, 1998.
[2] J.S. de Bruin, T.K. Cocx, W.A. Kosters, J. Laros and J.N. Kok, “Data
mining approaches to criminal career analysis,” in Proceedings of the Sixth
International Conference on Data Mining (ICDM’06), pp.171-177, 2006.
[3] J. Han and M. Kamber, “Data Mining: Concepts and Techniques,” Morgan
Kaufmann publications, pp. 1-39, 2006.
[4] J. Mena, “Investigative Data Mining for Security and Criminal Detection”,
Butterworth Heinemann Press,
pp. 15-16, 2003.
[5] S.V. Nath, “Crime pattern detection using data mining,” in Proceedings of
the 2006 IEEE/WIC/ACM InternationaConference on Web Intelligence and
Intelligent Agent Technology, pp. 41-44,2006.
[6] Brown, D.E and Hagen, S., “Data association methods with application to
law enforcement.” Decision Support Systems, 34, 2000, 369-378.
[7] Faith Ozgul, Claus Atzenbeck, AhmetCelik, Zeki, Erdem, “Incorporating
data Sources and Methodologies for Crime Data Mining,” IEEE proceedings,
2011.
[8] H.Chen, W.ChungXu, G.Wang, Y.Qin and M.Chen, “Crime Data Mining:
A General Framework and Some Examples,” computer, vol.37, 2004.
[9] H.Chen, W.Chung, Y.M.Chan, J.Xu, G.ang, R.Zheng and H. Atabakch,”
Crime Data Mining: An Overview and Case Studies,” in proceedings of the
annual national conference on digital government research, Boston, pp.1-5,
2003.
[10] M.Chan, J.Xu,and H.Chen,”Extracting Meaningful Entities from Police
Narrative Reports,” in proceedings of the National Conference on Digital
Government Research,pp.271-275,2002.
[11] R.Adderly and P.B. Musgrove, “Data Mining Case Study: Modeling the
behavior of offenders who commit sexual assaults,” in proceedings of the 2006
IEEE/WIC/ACM Conference on Web Intelligent Agent Technology, pp.41-44,
2006.
[12] Z.Huang,”Extensions to the K-means algorithm for clustering large
datasets with categorical values, “Data Mining and Knowledge Discovery,
vol.2, pp.283-304, 1998.
[13] L.Ma, Y.Chen, H.Huang, “AK-Modes: A weighted Clustering Algorithm
for Finding Similar Case Subsets,” 2010.
[14] Hao Cheng, Kien A. Hua and Khanh Vu, “Constrained Locally Weighted
Clustering,” journal proceedings of the VLDB Endowment, vol . 1, no .2, 2008
[15] Guenael cabanes and Younes bennani, “A Simultaneous Two-Level
Clustering Algorithm for Automatic Model Selection,” ICMLA '07
Proceedings of the Sixth International Conference on Machine Learning and
Applications, p p. 316-321, 2007
[16] Kilian Stoffel, Paul Cotofrei and Dong Han, “Fuzzy Methods for Forensic
Data Analysis,” European Journal of Scientific Research, Vol.52 No.4, 2011,
[17] Anuska Ferligoj, “Recent developments in cluster analysis,”
Telecommunication Systems, vol . 1,issue 4, 205–220, 2003

Survey of Data Mining Techniques on Crime Data Analysis

More Related Content

What's hot (19)

Similar to Survey of Data Mining Techniques on Crime Data Analysis (20)

More from ijdmtaiir (20)

Recently uploaded (20)

Survey of Data Mining Techniques on Crime Data Analysis