Pooja Bhagavat Memorial Mahajana PG centre,

KRS road, Metagalli, Mysuru-570016

" An Efficient Way to Detect Pickpockets from Automated Fare Collection Using Data Mining
Techniques"
Under the guidance of,
Presented by,
Mrs.Shobha.D Dharnish N C
VI Sem MCA
Associate Professor, Reg. No:18MCA058

DOS of Computer Science,

ABSTRACT
 
 Automated Fare Collection(AFC) is a collection of components that automate the
ticketing system of the public transportation network.
 Massive data collected by automated fare collection (AFC) systems provide opportunities
for studying both personal traveling behaviors and collective mobility patterns in urban
areas.
 Stopping pickpockets in the public transit system has been crucial for improving
passenger satisfaction and public safety. Even so, it is challenging to findout thieves
from regular passengers.
 In this project, I am developing a suspect detection and surveillance system, which can
identify pickpocket suspects based on their daily transit records.
INTRODUCTION
 The public transportation system passengers have been the major aim for pickpockets. In
several cities, thefts occurs regularly in public transportation systems, for the reason that
passengers tends to pay less attention to their things when they are in a rush or in a crowded
environment.
 It is difficult to provide a smart observation and tracking tool for the security personnel of the
transportation systems. The improving technologies like information technology and data
processing have capabilities; through transactional records collected by Automated Fare
Collection (AFC) systems.
 Actually, it is probable to identify thieves using AFC records because behavioral differences
are coined in the mobility footprints, which can help to separate suspects from regular
passengers.
 The initial challenge is how to recognize useful attributes to differentiate thieves from normal
passengers. These features should not only help us understand the behaviors of pickpockets,
but also help us build a suspect detection and tracking system for supporting the security
personnel.
 OBJECTIVES
 First, to identify useful features to distinguish thieves from regular passengers. These features should not only help
us understand the behaviors of pickpockets, but also help us build a suspect detection and tracking system for
supporting the security personnel.

 Second, using regular outlier detection methods tends to result in a large number of false positives. In particular,
not every trip made by a regular passenger looks normal.

 Third, a large number of AFC records are being collected from millions of passengers, only a tiny fraction of which
are pickpockets. Identifying such a small group of people in such a large-scale dataset is like looking for a needle
in the haystack.

 Finally,also need to effectively transform our knowledge based on model development into a decision support
system, so that real-time, personalized deployment recommendations could be made to help to guide security
personnel to perform their work more efficiently.
PROBLEM STATEMENT
 With rapid advances in information technology and infrastructure, transactional
records collected by Automated Fare Collection (AFC) systems.

 Most existing studies focus on identifying regular, collective mobility patterns, such
as commute flows and transit networks.

 It is possible to detect thieves using AFC records because behavioral differences

logged in the mobility footprints may be used to separate suspects from regular
passengers.

 Designing an intelligent system that automatically extracts specific, identified

behavioral features and dynamically detects and tracks pickpocket suspects has
become a possibility.
 EXISTING SYSTEM
 The Existing System focuses on finding patterns in passenger activity records. Such
knowledge can be useful in a variety of applications, and plays a vital role in effectively
finding and satisfying passenger needs.
 Existing studies that detect anomalies in urban sensing data can be divided into two
categories: those based on locations, and those on trajectories

DISADVANTAGES
 There is no Smart Card Based Travelling due to Only Manual Passengers Activity Patterns
and Manual Transit Records.
 Since thieves involved in victim reported events were not caught, the system could not
identify them according to the no-checkout rule.
 PROPOSED SYSTEM
 In the proposed system, the system adopted a comprehensive approach to the pickpocket
detection problem. The overall framework of our solution is illustrated in this system.
 Firstly, we identified a number of features that may be extracted from AFC records
and are potentially useful for distinguishing thieves from regular passengers.
 Most importantly, a real system for the end user was designed and tested using real-
world, large-scale data.

ADVANATAGES
 Effective techniques for Abnormal Traveling Behavior Detection.

 Smart Card Based Transit Record Maintenance.

 H/W AND S/W REQUIREMENTS

Hardware Requirements Software Requirements

 Processor - Pentium –IV Operating System - Windows 7
 RAM - 4 GB (min) Coding Language - Java/J2EE(JSP,Servlet)
 Hard Disk - 20 GB S
 Key Board - Standard Windows
Front End - J2EE
 Mouse - Two or Three Button
 Monitor - SVGA Back End - MySQL
IMPLEMENTATION
SYSTEM ARCHITECTURE
DATA FLOW DIAGRAM
USE CASE DIAGRAM
SCREENSHOTS
 CONCLUSION

In this project, developed a suspect detection and tracking system by mining large-scale
transit records. The system assists in identifying pickpocket suspects and enables active
surveillance in high-risk areas. Specifically, first constructed a feature representation
for profiling passengers. Then, established a two-step framework to distinguish regular
passengers from pickpocket suspects. Finally, support the real-world datasets from
multiple sources for model training and validation, and implemented a prototype system
for end users.
FUTURE ENHANCEMENT

 In future we can implement the same concepts on the live server or cloud.

 In future we can also implement this concepts for Andriod and Ios smartphones.