Crime

analysisProcedu
re

International Conference on Innovation and Advance Technologies in Engineering 1 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm
PROCEDURE
Crime analysis is an analytic process which provides relevant information about various crime
patterns. This project forms clusters using K means algorithm which plays a vital role in
analysing and predicting crimes. It clusters various types of crimes by accessing the stored
data. It also helps in identifying relevant crime patterns, hidden links, link prediction and
statistical analysis of crime data. The prime goal of this study is to examine clustered analysis
as an innovative way to detect and help reduce crime and safeguard society.
It is vital to analyse crime due to following reasons:
 To inform law enforcers about specific crime trends promptly
 Analyse crime from information existing injustice system and public domain.
 Predict crimes based on existing data and anticipation of using different data mining
techniques.

Data Collection and Classification

Police records contain a large amount of crime data. This data is made accessible by
the National Crime Bureau of Records. The data we used is recorded throughout the country
over the years. The genuine cases are not filtered and therefore may contain some incorrect
values. Processing has to be done for accurate information. According to their characteristics,
the data set is categorised into different groups. Grouping is majorly done based on state or
city where various types of crimes occur. K-mean algorithm is used to cluster data.

Pre-processing
Data pre-processing is a data mining technique that involves transforming raw data into an
understandable format. Real-world data is often incomplete, inconsistent, and lacking in certain behaviours or
trends, and is likely to contain many errors. In this step, we clean the data and make it ready for our clustering
algorithm so that it is easy to make predictions and reduce any errors.

Classification
Classification is a step-in data mining that allocates target categories or classes to the items in the collection. We
can correctly predict the destination class for each case in the data. In this case, a classification model is used to
group crimes based on various parameters.
For example, Crimes can be classified based on severity - Low Risk, High risk.

Clustering
It is used to determine association rules which highlight general trends in the database. In this step, we
identify the crime pattern for a particular place and correspond to each location we take the attributes and
predict some pattern to that particular cluster. When a new case comes, it follows the same pattern and says that
this area has a chance for crime occurrence.

Crime prediction
It is simple to understand and interpret and it works well with large datasets.

International Conference on Innovation and Advance Technologies in Engineering 2 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm

Cluster
K means cluster analysis involves tracking crime rate changes throughout the years, which is further used in
data mining to project those changes into the future. Here we consider homicide crime and plot it with year and
analysis variation in a graph on cluster formed.

Fig 2 (a) Non Classified Clusters (b) Classified Clusters

International Conference on Innovation and Advance Technologies in Engineering 3 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm

Visualization
Security is an aspect which must be given higher priority by all political and government organisations
worldwide, aiming to reduce crime incidences. As data mining is the appropriate field to apply on more massive
datasets, the knowledge gained from data mining approaches will be highly useful and will support the police
force. So,In this paper crime analysis is done by performing k-means clustering on crime dataset.

Fig 3 Clustered and Unclustered Data

International Conference on Innovation and Advance Technologies in Engineering 4 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm
Implementation Plan

As we need to perform unsupervised learning on unlabelled data, we use K-Means clustering; The
variable K represents the number of groups. Based on the features provided in the data, the algorithm works
iteratively to assign each data point to one of the K groups. Data points are clustered based on feature similarity.
The results of the K-means

The clustering algorithm is

1. The centroids of the K clusters, which can be used to label new data
2. Labels for the training data (each data point is assigned to a single cluster)
Rather than defining groups before looking at the data, clustering allows you to find and analyse the groups that
have formed organically.

International Conference on Innovation and Advance Technologies in Engineering 5 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm
Table 1 - Database Example Table
ID Integer

Crime Char

Location Char

Time Datetime

Occasion (if any) Varchar

Purpose Varchar

Accused Char

International Conference on Innovation and Advance Technologies in Engineering 6 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm

Future Scope
As of now, the project relies on manual input from a human (a police officer) in order to enter details in
the database. If we can make this a centralised system and connect it to all the police stations countrywide and
make FIR reporting digital, then it would be quite easier to predict crimes in that particular location and
recognise patterns in them. It would also encourage citizens to track their E-FIR online. We can also avoid
corruption as the government can keep a track on the number of cases registered and their solvability rate which
can help them utilise their resources better.

International Conference on Innovation and Advance Technologies in Engineering 7 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm

Conclusion
In this system, we get to classify and cluster to improve the accuracy of location and pattern-based
crimes. This software predicts frequently occurring crimes and its locations, especially for particular state,
region and occasions. Also, time is an essential factor in the occurrence of crimes, and we shall predict the time
as well in this application.

International Conference on Innovation and Advance Technologies in Engineering 8 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai
 Crime Rate Prediction Using K-Means Algorithm

References
[1]. Detection Analysis & Prediction Meet Meet Timbadia, Ajit Yadav Information Technology, University Of Mumbai, Shree
L.RTiwari College Of Engineering,Thane, India, Year – 2017.
[2]. Z-crime: A Data Mining Tool for The Detection of Suspicious Criminal Activity Based on Decision Tree Mugdha Sharma
Information Technology, University of Mumbai, Shree L.R Tiwari College of Engineering, Thane, India, Year – 2015.
[3]. A Survey of Data Mining Techniques for Analyzing Crime Patterns Ub on Than sataomwatan. Second Asian conference on defence
technology. Year -2017.
[4]. Crime Prediction and Forecasting in Tamilnadu Using Clustering Approaches S Sivaranjani, S Sivakumari, M Aasha. Dept. Of
information technology. University of Mumbai L.R. Tiwari College Year- 2017.
[5]. K. Bogahawatte, and S. Adikari, “Intelligent criminal identification system,” IEEE 2013 the 8th International Conference on
Computer Science & Education (ICCSE) 2013 April 26-28, 2013. Colombo, Sri Lanka.

International Conference on Innovation and Advance Technologies in Engineering 9 | Page

Atharva College of Engineering Malad Marve Road, Charkop Naka, Malad West Mumbai