A CROSS METHODOLOGY FOR CREDIT CARD

DECEPTION RECOGNITION USING ROUGH SET

AND DECISION TREE TECHNIQUE

UNDER THE GUIDENCE OF

N.R.G.K. PRASAD I.ANUSHA (315232915004)
Asst. Professor, M.Tech,(Ph.D) M.Tech, (CST)
 ABSTRACT
The use of credit cards is prevalent in modern day society.
The number of credit card fraud cases is constantly increasing in spite
of the chip cards worldwide integration and existing protection
systems.
This is why the problem of fraud detection is very important now
In this paper the general description of the developed fraud detection
system and comparisons between models based on using of decision
tree rough set and cryptography techniques which can be used in credit
card fraud detection mechanisms.
 INTRODUCTION

In the credit card transaction credit card credentials for online transaction
and physical card for offline transaction is used. When physical card is
given for the transaction, the credit card details or the card itself can be
stolen and if the owner of the card is not aware about the loss of card it may
result in the fraudulent transactions by unauthorized users and the financial
company is at the loss.
 EXISTING SYSTEM
Credit Card is plastic money and is widely used as a mode of payment.
Credit card owners are increasing at a high rate and all users transact with a
sense of security and confidence, hence credit card
security is a prime concern for online transactions.
It is important to safeguard the process so that the e-commerce platform
providers as well as the credit card
providers along with the user using the credit card are not at a loss. Secure
services can be provided only with a reliable and safe model to protect the
transactions
 PROPOSED SYSTEM
This section deals with the proposed work. This work is meant for improving the
credit card fraud detection.
This proposed work is made up of various important facts. These are mentioned as
follows:
Preprocessing using Rough Set
Rough sets theory was introduced as a mathematical tool for data analysis.
Data preprocessing reduces the data complexity and offers better chances for
subsequent analysis.
The best features can be found by determining the dependency between any
conditional feature and the decision feature..
PROPOSED ARCHITECTURE

Apply j48

Reduced data set

Apply Rough set

Apply data set

Data Flow Diagram
 NUMBER OF MODULES
The modules involved are
Admin/User Module
Purchase & Billing Module
Update Module
Deposit Module

Admin/User Module
In our proposed system provide the login credentials for every user so every user can login to our
system by using their credentials but Admin can create the new user only. Every user credentials
storing and retrieving based cryptography system while creating the new user and login user.

Purchase & Billing Module

This module is used for selecting the items from stock which items they want and generate the bill.
After generating the bill if user wants to pay the money by using credit cards then system generate
the OTP based on decision tree technique, by using this OTP money transaction performed.

Update Module
In this module user can modify or update their profile based on user ID and security questions which
are mentioned while user registration. Main use of this functionality is modified their credentials and
security questions, answers while hacking their credentials by hackers.

Deposit Module
This system provides the option for deposit the money into our account this module used for pay
the amount for credit cards.
 MINIMUM HARDWARE REQUIREMENTS

System : Dual core processor

Hard Disk : 80 GB.
RAM : 1 GB
 MINIMUM SOFTWARE REQUIREMENTS

Operating system : Windows 7 Ultimate (32-bit) OS

Front End : Visual Studio 2010
Coding Language : C#.NET
Database : SQL Server 2008
 USE CASE DIAGRAM:
System

Registration

User
Login

Add/View Products

Administration
View Orders

Data Sets

Dataset Keys for

Verification

Success

Logout
 CLASS DIAGRAM

webserver
admin users
mobile users
users
groups
feed backs
add_details()
ranks
view_details()
friend match graph
view_all_users()
register
add groups()
view frauds()
add details()
logout()
view details()
logout()

login
user name register
password
name
password
login()
email
register()
mobile
reset()
end user uid and password
user name
cards register()
reset()
view profile()
transaction()
failed tans()
logout()
 SEQUENCE DIAGRAM

web server user

admin

add/view details search products

add/view card information

check transaction

generate keys
security key

verification

verify keys
 ACTIVITY DIAGRAM

start

user register

yes login no

add data no
user name and
password wrong
log out
view products

buy product

yes
privacy data
sets keys

verify keys

save
transactions
SCREEN SHOTS
USER REGISTRATION
 UPDATE USER PROFILE

In this page user can modify or update their profile based on user Id and
security questions which are mentioned while user registration.
Main use of this functionality is modified their credentials and
security questions, answers while hacking their credentials by hackers.
USER PROFILE UP DATION
 ADD PRODUCTS

This functionality was used for adding the new products which their want
to adding, after adding the items or products users can purchase those
products by using their credit cards.
admin can only add the products then users can purchase those
product by their using credit cards
ADD PRODUCTS
VIEW PRODUCTS
PURCHASE BILL
TRANSACTION
List of users
 CONCLUSION
Credit card use is prevalent and used by everyone. Owing to the internet
platform payment system is vulnerable to fraud attacks. The fraud detection
systems in credit card transactions need to be very robust and precise, giving
minimum false alerts and exactly classifying the fraud and non-fraud
transactions. The system proposed in this article is very much effective and
accurate. Last section showed that the performance of the proposed system is
much better than existing work. Using this framework, financial institutions
can compare the transaction information with the historical profile patterns to
predict the probability of being fraudulent for a new transaction, and provide
a scientific basis for the authorization mechanisms. Furthermore, resources of
the institutions can be focused on more suspicious transactions to decrease
the fraud levels.