crime pattern detection using k-means clustering

43
CRIME PATTERN DETECTION

Upload: reuben-george

Post on 14-Jun-2015

826 views

Category:

Documents


4 download

TRANSCRIPT

Page 1: Crime Pattern Detection using K-Means Clustering

CRIME PATTERN DETECTION

Page 2: Crime Pattern Detection using K-Means Clustering

CONTENTS

• What is crime?• Types of crime• What makes one

commit crime?• Statistics• Effects of crime

• Steps of Crime Pattern Detection• Clustering • Pattern Analysis• Pattern Results

• Advantages of CPD• Limitations of CPD• Conclusions• Future Direction

Page 3: Crime Pattern Detection using K-Means Clustering

CRIMEDEFINITION

“an action which constitutes a serious offence against an individual or the state and is punishable by law.”

- Concise Oxford Dictionary

“an act or the commission of an act that is forbidden or the omission of a duty that is commanded by a public law and that makes the offender

liable to punishment by that law” - Merriam Webster Dictionary

Page 4: Crime Pattern Detection using K-Means Clustering

TYPES OF CRIMEORGANIZED

CRIME

• Drug trafficking• Gunrunning• Money laundering• Extortion• Murder for hire• Fraud• Human trafficking• Poaching

PROPERTY CRIME

• Burglary• Theft• Motor vehicle theft• Arson

CORRUPTION

Page 5: Crime Pattern Detection using K-Means Clustering

WHAT MAKES ONE COMMIT CRIME?

Page 6: Crime Pattern Detection using K-Means Clustering

• Peer pressure• Criminals have not been taught the difference

between ‘right and wrong.’ • Mental illness. • A failure to rehabilitate ex-offenders back into

society

Page 7: Crime Pattern Detection using K-Means Clustering

The sociologist Zygmunt Bauman argues that “criminals steal status items in order to appear

‘normal’ within such a materialistic society”

Page 8: Crime Pattern Detection using K-Means Clustering

THE PEAK AGE OF CRIMINAL ACTIVITY IS DURING

THE YEARS 16-25.

WHAT MAKES THEM COMMIT THEM?

Page 9: Crime Pattern Detection using K-Means Clustering

• Boys often have to ‘prove’ their masculinity which can, at times, result in criminal activity

• The likelihood of a young person belonging to a subculture is high, and some subcultures engage in criminal behavior

• Young people may have few legitimate means available of acquiring material goods

• Less responsibilities• Teenage rebellion can lead to people breaking

the law

Page 10: Crime Pattern Detection using K-Means Clustering

NEGATIVE IMPACTS OF CRIME UPON AN AREA

Page 11: Crime Pattern Detection using K-Means Clustering

• Depopulation, particularly in urban areas• High levels of crime may damage community

spirit and result in less neighborliness. • High crime levels can contribute to

environmental poverty• Once a region with a high level of crime is

labeling as a bad area, it might become a ghetto

Page 12: Crime Pattern Detection using K-Means Clustering

SEVERAL CAUSES OF DEVIANT BEHAVIOR THAT YOU ALSO NEED TO BE AWARE OF

Page 13: Crime Pattern Detection using K-Means Clustering

• People may feel alienated from society. • Deviant behavior may simply be the product

of teenage rebellion• In order to conform to the subculture of that

group, people adopt the ways of the subculture.

Page 14: Crime Pattern Detection using K-Means Clustering

STATISTICS

Page 15: Crime Pattern Detection using K-Means Clustering

YEARTOTAL COG.

CRIMES UNDER IPC

MURDER KIDNAPPING ROBBERY BURGLARY RIOTS

1953 6,01,964 9,802 5,261 8,407 147,379 20,529

2006 18,78,293 32,481 23,991 18,456 91,666 56,641

% Change in 2006

over 1953212.0 231.0 356.0 120.0 -38.0 176.0

Page 16: Crime Pattern Detection using K-Means Clustering

CRIME PATTERN DETECTION

Page 17: Crime Pattern Detection using K-Means Clustering

Questions investigators face• Are there correlations between the crime type and the

location of the incident? • What are the distributions of crime types involving suspects

of different ethnic origin? • How can I quickly extract reports characterized by certain

parameters of interest? – For example: robberies performed by white teenagers

involving the knife threat.• Are there correlations between the type of crime, weapon

employed, and the location of the incident?• What is the most typical weapon in cases when high school

students are charged with weapon possession?

Page 18: Crime Pattern Detection using K-Means Clustering

Why crime pattern analysis?

To implement a data analysis framework which works with the geospatial plot of crime and helps to improve the productivity of the detectives and other

law enforcement officers.

To use semi-supervised learning technique here for knowledge discovery from the crime records and to

help increase the predictive accuracy.

Page 19: Crime Pattern Detection using K-Means Clustering

Steps involved in crime pattern analysis

Determine geo-spatial plots of crime

in a city

Using proper clustering techniques to identify patterns

Analyzing patterns and drawing conclusions

Crime Type

Suspect Race

Suspect Gender

Suspect Age group

Victim Age group

Weapon

Robbery B M Middle Elderly Knife

Robbery W M Young Middle Bat

Robbery B M ? Elderly Knife

Robbery B F Middle Young Piston

Page 20: Crime Pattern Detection using K-Means Clustering

STEP #1

DETERMINE GEO-SPATIAL PLOTS OF CRIME IN A CITY1. Collecting Information

1. Police department records2. Electronic systems for crime reporting. (N.D.A)3. Narrative or description of the crime4. Modus Operandi

2. Translate occurrences of crime into plots on a geographical map of a city

Page 21: Crime Pattern Detection using K-Means Clustering

STEP #2

USING PROPER CLUSTERING TECHNIQUES TO IDENTIFY PATTERNS

Page 22: Crime Pattern Detection using K-Means Clustering

CLUSTERING

Data mining terminology a cluster is group of similar data points (a possible crime pattern)

Crime terminology a cluster is a group of crimes in a

geographical region or a hot spot of crime.

Page 23: Crime Pattern Detection using K-Means Clustering

Cluster analysis or clustering is the assignment of a set of observations into subsets (called clusters) so that observations in the same cluster are similar in some sense.

Page 24: Crime Pattern Detection using K-Means Clustering

Clustering is a method of unsupervised learning, and a common technique for statistical data analysis used in many fields, including 1. machine learning, 2. data mining, 3. pattern recognition, 4. image analysis, 5. information retrieval, and 6. bioinformatics.

Page 25: Crime Pattern Detection using K-Means Clustering

Clustering Technique

Task of identifying groups of records that are similar between themselves but different from the rest of the data and of finding the variables providing the best clustering

Clusters will useful for identifying a crime spree committed by one or same group of suspects.

Page 26: Crime Pattern Detection using K-Means Clustering

These clusters will then be presented to the detectives to drill down using their domain expertise.

Automated detection of crime patterns, allows the detectives to focus on1. crime sprees first and solving one of these crimes results

in solving the whole spree” 2. groups of incidents suspected to be one spree, the

complete evidence can be built from the different bits of information from each of the crime incidents.

Page 27: Crime Pattern Detection using K-Means Clustering

Why Clustering?

• Crimes vary in nature widely • Nature of crimes change over time• Crime database often contains several

unsolved crimes.• Less predictive quality for solving future

crimes

Page 28: Crime Pattern Detection using K-Means Clustering

Why Clustering?

In order to be able to detect newer and unknown patterns in future, clustering

techniques work better.

K-Means Clustering was used here.

Page 29: Crime Pattern Detection using K-Means Clustering

K-Means Clustering

The k-means algorithm assigns each point to the cluster whose centroid is nearest. The center is the average of all the points in the cluster

Example: The data set has three dimensions and the cluster has two points: X = (x1,x2,x3) and Y = (y1,y2,y3). Then the centroid Z becomes Z = (z1,z2,z3) , where , and

Page 30: Crime Pattern Detection using K-Means Clustering

K-Means Algorithm

1. Choose the number of clusters, k.2. Randomly generate k clusters and determine the cluster

centers, or directly generate k random points as cluster centers.

3. Assign each point to the nearest cluster center, where "nearest" is defined with respect to one of the distance measures discussed above.

4. Recompute the new cluster centers.5. Repeat the two previous steps until some convergence

criterion is met (usually that the assignment hasn't changed).

Page 31: Crime Pattern Detection using K-Means Clustering

STEP 1 STEP 2

STEP 3 STEP 4

Page 32: Crime Pattern Detection using K-Means Clustering

WHY DATA MINING APPROACH?1. not be easy for a computer

data analyst or detective to identify these patterns by simple querying

2. deal with enormous amounts of data and dealing with noisy or missing data about the crime incidents

Page 33: Crime Pattern Detection using K-Means Clustering

STEPS INVOLVED IN CLUSTERING

1. Sorting of records – first sort will be on the most important characteristic based on the detective’s experience.

Crime Type Suspect Race

Suspect Gender

Suspect Age group

Victim Age group

Weapon

Robbery B M Middle Elderly Knife

Robbery W M Young Middle Bat

Robbery B M ? Elderly Knife

Robbery B F Middle Young Piston

Page 34: Crime Pattern Detection using K-Means Clustering

2. Use data mining to detect much more complex patterns since in real life there are many attributes or factors for crime and often there is partial information available about the crime.

3. Identify the significant attributes for the clustering.

4. Placing different weights on different attributes dynamically based on the crime types being clustered

Page 35: Crime Pattern Detection using K-Means Clustering

5. Cluster the dataset for crime patterns and then present the results to the detective or the domain expert along with the statistics of the important attributes.

6. The detective looks at the clusters, smallest clusters first and then gives the expert recommendations.

7. unsolved crimes can be clustered based on the significant attributes and the result is given to detectives for inspection

Page 36: Crime Pattern Detection using K-Means Clustering

STEP #3ANALYZING PATTERNS AND DRAWING

CONCLUSIONS

Page 37: Crime Pattern Detection using K-Means Clustering

PATTERN RESULTS

Page 38: Crime Pattern Detection using K-Means Clustering
Page 39: Crime Pattern Detection using K-Means Clustering

ADVANTAGES OF CRIME PATTERN DETECTION

Page 40: Crime Pattern Detection using K-Means Clustering

1. Learn from historical crime patterns and enhance crime resolution rate.

2. Preempt future incidents by putting in place preventive mechanisms based on observed patterns.

3. Reduce the training time for officers assigned to a new location and having no prior knowledge of site-specific crime patterns.

4. Increase operational efficiency by optimally redeploying limited resources (like personnel, equipment, etc.) to the right place at the right time.

Page 41: Crime Pattern Detection using K-Means Clustering

LIMITATIONS OF CRIME PATTERN DETECTION

Page 42: Crime Pattern Detection using K-Means Clustering

1. Crime pattern analysis can only help the detective, not replace them

2. Data mining is sensitive to quality of input data that may be inaccurate, have missing information, be data entry error prone

3. Mapping real data to data mining attributes is not always an easy task and often requires skilled data miner and crime data analyst with good domain knowledge

Page 43: Crime Pattern Detection using K-Means Clustering