International Journal of Computer Applications (0975 – 8887)
Volume 161 – No 13, March 2017
6
A Novel Idea for Credit Card Fraud Detection using
Decision Tree
Prajal Save
St. John College of
Engineering and
Management
Palghar
Pranali Tiwarekar
St. John College of
Engineering and
Management
Palghar
Ketan N. Jain
St. John College of
Engineering and
Management
Palghar
Neha Mahyavanshi
St. John College of
Engineering and
Management
Palghar
ABSTRACT
Online shopping and banking has increased by the growth of
internet and by use of credit card. Along with this number of
credit card fraud is also increased. Many modern techniques
based on Artificial Intelligence, Data warehousing has
evolved in detecting various credit card fraudulent
transactions. We proposed a system which detect fraud in
credit card transaction processing using a decision tree with
combination of Luhn's algorithm and Hunt's algorithm.
Luhn’s algorithm is used to validate the card number. Address
matching rule checks whether the Billing Address and
Shipping Address match or not. This check does not guarantee
whether a transaction is fraud or genuine. But if the two
addresses match, the transaction can be classified as genuine
with a high probability. Else, the transaction is labelled as
suspect. A customer usually carries out similar types of
transactions in terms of amount, which can be visualized as
part of a cluster. Since a fraudster is likely to differ from the
customer’s account, his transactions can be detected as
exceptions to the cluster – a process known as outlier
detection.
General Terms
Credit card fraud, online Transaction, Electronic Commerce
Keywords
Electronic Commerce, Credit card fraud, address matching,
spending pattern, Luhn's Algorithm, Outlier Detection,
Heuristic function, Bayes Theorem.
1. INTRODUCTION
With rapid advancement of e-commerce, use of credit cards
for purchases has exponentially increased. Unfortunately,
fraudulent use of credit cards has also become a source of
crime. Credit card fraud is a most popular term for theft and
fraud committed using or involving a payment card, such as a
credit card or debit card, as a fraudulent source of funds in a
transaction. Credit card fraud is also an appendage to identity
theft. According to the United States Federal Trade
Commission, while identity theft had been holding steady for
the last few years, it saw a 21 percent increase in 2008.
However, credit card fraud, that crime which most people’s
privilege with ID theft, decreased as a percentage of all ID
theft complaints for the sixth year in a row.Financial
institutions employ various fraud prevention models for
tackling this problem. But fraudsters are adaptive, and given
time, they devise several ways to intrude such protective
models. Despite the best efforts of the financial institutions,
law enforcement agencies and the government, credit card
fraud continues to rise. Fraudsters nowadays may constitute of
a very inventive, intellect and fast moving fraternity. Several
techniques for the detection of credit card fraud have been
proposed in the last few years.
2. LITERATURE SURVEY
Syeda et al. [1] have suggested the use of parallel granular
neural networks for speeding up the data mining and
knowledge discovery process. Maes et al. [2] have outlined an
automated credit card fraud detection system by Artificial
Neural Network - ANN as well as Bayesian belief networks -
BBN. They show that BBN gives better results related to
fraud detection and the training period is faster whereas the
actual detection process is substantially faster with ANN. The
neural network based methods are, in general, fast but not so
accurate an retraining the neural networks is quite taxing.
Chen et al. [3] propose a method in which an online
questionnaire is used to collect questionnaire-responded
transaction (QRT) data of users. Further it uses a support
vector machine (SVM) trained with this data and the QRT
models are used to predict new transactions. Chen et al. [4]
have recently presented a personalized approach for credit
card fraud detection that employs both SVM and ANN. It tries
to prevent fraud for users even without any transaction data.
However, these systems are not fully automated and depend
on the user‟s expertise level.
Chan et al. [5] divide a large set of transactions into smaller
subsets and then apply distributed data mining for building
models of user behavior. The resultant base models are then
combined to generate a meta-classifier for improving
detection accuracy. Brause et al. [6] have explored the
possibility of combining advanced data mining techniques and
neural networks to obtain high fraud coverage along with a
low false alarm rate. Use of data mining is also developed by
Chiu and Tsai [7] . They consider web services for data
exchange among banks. A fraud pattern mining (FPM)
algorithm has been elaborated for mining fraud association
rules which give information regarding the features that exist
in fraud transactions. Banks enhance their original fraud
detection systems by using the new fraud patterns to halt
attacks. While data mining techniques are relatively accurate,
they are inherently slow.
Aleskerov et al. [8] present CARDWATCH, a database
mining system used for credit card fraud detection. The
system, based on a neural learning module, provides an
interface to a variety of commercial databases. Kim and Kim
have identified skewed distribution of data and combination
of legitimate and fraudulent transactions as the two main
reasons for the complexity of credit card fraud detection [9] .
Based on this observation, they use fraud density of real
transaction data as a confidence value and generate the
International Journal of Computer Applications (0975 – 8887)
Volume 161 – No 13, March 2017
7
weighted fraud score to reduce the number of wrong
detections
3. RELATED WORK
Various existing fraud techniques majorly explore decision
trees, genetic algorithms, clustering techniques and neural
networks. Recently there has been increased in use of parallel
granular neural networks for accelerating the data mining and
knowledge discovery methods. It seems that BBN (Bayesian
Belief networks) gives better results related to fraud detection
and the training period is faster whereas the actual detection
process is substantially faster with ANN. The neural network
based methods are, in general, fast but not so accurate and
retraining the neural networks is quite taxing.
Genetic algorithms was first introduced by Holland(1975).
Genetic algorithms obtain better solutions as time progresses.
Fraud detection problem is classification problem, in which
some of statistical methods many data mining algorithms have
proposed to solve it. Although decision trees are more
popular. Fraud detection has been usually in domain of E-
commerce, data mining [10]. GA is used in data mining most
probably for variable selection [11] and is mostly coupled
with other DM algorithms [12]. And their combination with
other techniques has a very good performance. GA has been
used in credit card fraud detection for reducing the wrongly
classified number of transactions [12]. And is easy accessible
for computer programming language implementation, thus,
make it powerful in credit card fraud detection. But this
method has high performance and is quite expensive.
A Hidden Markov Model is a double embedded stochastic
process used to model much more complicated stochastic
processes as compared to a traditional Markov model. If an
incoming credit card transaction is denied by the trained
Hidden Markov Model with sufficiently high probability, it is
considered to be fraudulent transactions. HMM[13], Baum
Welch algorithm is used for training purpose and K-means
algorithm for clustering purpose. HMM stores data in the
form of clusters depending on three price value ranges low,
medium and high[14]. The probabilities of initial set of
transaction have chosen and Fraud Detection System checks
whether transaction is genuine or fraudulent. Since HMM
maintains a log for transactions it reduces tedious work of
employee but produces high false alarm as well as high false
positive[15]
Neural networks have been generally used in fraud detection.
Neural network is a set of Connected input/output units and
each connection has a weight present with it. During the
learning phase, network learns by adjusting weights to guess
the correct class labels. Fraud detection methods based on
neural network are the most popular ones. An artificial neural
network [16][17] consists of an interconnected group of
artificial neurons. The principle of neural network is inspired
by the functions of the brain especially pattern recognition
and associative memory [18]. The neural network diagnose
similar patterns, predicts future values or events based upon
the associative memory of the patterns it was learned. It is
generally applied in classification and clustering. The
advantages of neural networks over other techniques are that
these models are able to learn from the past and thus, enhance
results as time passes. They can also extract rules and predict
future activity based on the current situation. By employing
neural networks, effectively, banks can detect fraudulent use
of a card, faster and more efficiently. Through the reported
credit card fraud studies most have focused on using neural
networks. In more practical terms neural networks are non-
linear statistical data modeling tools. They can be used to
design complex relationships between inputs and outputs or to
find patterns in data
4. PROPOSED SYSTEM
The proposed model comprised of six steps. Firstly, Luhn's
Test is used to validate card numbers. Then, two rules ie.
Address Mismatch and Degree of Outlierness are used to
analyze the deviation of each incoming transaction from the
normal profile of cardholder. These two steps compute initial
beliefs. The initial belief values are combined to obtain an
overall belief by applying Advanced Combination Heuristic in
step four. Step five looks into the spending history to extract
characteristic information about genuine and fraud
transactions. The overall belief is further strengthened or
weakened in the final step using Bayes‟ Theorem, followed
by recombination of the calculated probability with initial
belief of fraud using advanced combination heuristic
Fig 1: Architecture Diagram
Card Number Validation
Luhn's Algorithm is used to validate card numbers that
distinguishing valid numbers from mistyped or otherwise
incorrect numbers.
Following standard algorithm is used to validate credit card
numbers, [14]
1. Reverse the order of the digits in the number.
2. Take the first, third, ... and every other odd digit in the
reversed digits and sum them to form the partial sum S1.
3. Taking the second, fourth and every other even digit in
the reversed digits. Multiply each digit by two and sum
International Journal of Computer Applications (0975 – 8887)
Volume 161 – No 13, March 2017
8
the digits if the answer is greater than nine to form partial
sums for the even digits.
4. Sum the partial sums of the even digits to form S2.
If S1+S2 ends in zero,then the original number is in the form
of a valid credit card number as verified by the
Luhn test.
For example, if the trial number is 49927398716,
1. Reverse the digits:
61789372994
2. Sum the odd digits:
6 + 7 + 9 + 7 + 9 + 4 = 42 = s1
3. The even digits:
1, 8, 3, 2, 9
Two times each even digit:
2, 16, 6, 4, 18
Sum the digits of each multiplication:
2, 7, 6, 4, 9
4. Sum the last:
2 + 7 + 6 + 4 + 9 = 28 = s2
5. S1+S2 = 70 which ends in zero which means that
49927398716 passes the Luhn's test.
Address Verification
This step is used for comparing Billing Address with Shipping
Address and the check is whether it matches or not. This
check does not guarantee whether a transaction is fraud or
genuine. But if the two addresses match,t he transaction can
be classified as genuine with a high probability. Else, the
transaction is labeled as suspect.
Outlier Detection
We have used DBSCAN (Density Based Spatial Clustering of
Application with Noise) to generate clusters, using transaction
amount as attribute. Any incoming transaction amount, that
does not belong to any cluster is detected as fraudulent. These
two steps compute initial belief.
Advanced Combination Heuristic Function
The initial belief values are combined to obtain an overall
belief.
Spending History Databases
It comprises of genuine Transaction Record(for individual
customers from their past behaviour) and Fraud Transaction
Record(from different types of past fraud data).We represent
each history transaction by set of attributes containing
information like card number, transaction amount and time
since last purchase. to extract characteristic information about
genuine and fraud transactions.
Bayes Theorem
The idea of belief revision is that, whenever new information
becomes available, it may require updating of prior beliefs.
Bayes Theorem theorem expresses how a subjective degree of
belief should rationally change to account for availability of
related evidence.
5. RESULT AND ANALYSIS
Credit Card verification is done by using Luhn’s algorithm
which checks whether the entered card number is valid or not.
The remaining phases of the proposed system will be
implemented in future work.
Input to the system is credit card number which performs
luhn’s algorithm process as explained in credit card validation
step .The output for this shows whether it passes the luhn’s
test or not.
6. CONCLUSION
In this paper we have brief discussion on credit card fraud
detection. Here we have shown how the system detect
whether an incoming transaction is fraud or genuine.In our
proposed model, we have found out validation of card are
genuine and very low false alarm.The relative studies and our
results sure that the correctness and effectiveness of the
proposed system is secure.
7. REFERENCES
[1] M. Syeda, Y.Q. Zhang, Y. Pan, “Parallel granular neural
networks for fast credit card fraud detection”,
Proceedings of the IEEE International Conference on
Fuzzy Systems, 2002, pp. 572–577.
[2] S. Maes, K. Tuyls, B. Vanschoenwinkel, B. Manderick,
“Credit card fraud detection using Bayesian and neural
networks”, Proceedings of the First International NAISO
Congress on Neuro Fuzzy Technologies, 2002
[3] R.C. Chen, M.L. Chiu, Y.L. Huang, L.T. Chen,
“Detecting credit card fraud by using questionnaire-
responded transaction model based on support vector
machines”, Proceedings of the Fifth International
Conference on Intelligent Data Engineering and
Automated Learning, vol. 3177, October 2004, pp. 800–
806.
[4] R.C. Chen, S.T. Luo, X. Liang, V.C.S. Lee,
“Personalized approach based on SVM and ANN for
detecting credit card fraud”, Proceedings of the IEEE
International Conference on Neural Networks and Brain,
October 2005, pp.810–815.
[5] P.K. Chan, W. Fan, A.L. Prodromidis, S.J. Stolfo,
“Distributed data mining in credit card fraud detection”,
Proceedings of the IEEE Intelligent Systems, 1999,
pp.67–74.
[6] R. Brause, T. Langsdorf, M. Hepp, “Neural data mining for
credit card fraud detection”, Proceedings of the International
Conference on Tools with Artificial Intelligence, 1999, pp. 103–
106.
[7] C. Chiu, C. Tsai, “A web services-based collaborative scheme
for credit card fraud detection”, Proceedings of the IEEE
International Conference on e-Technology, e-Commerce and e-
Service, 2004, pp. 177–181.
[8] E. Aleskerov, B. Freisleben, and B. Rao, “CARDWATCH: A
Neural Network Based Database Mining System for Credit Card
Fraud Detection”, Computational Intelligence for Financial
Eng., pp. 220-226, 1997.
[9] M.J. Kim and T.S. Kim, “A Neural Classifier with Fraud
Density Map for Effective Credit Card Fraud Detection,” Proc.
Int‟l Conf.Intelligent Data Eng. and Automated Learning, pp.
378-383, 2002.
[10] K.RamaKalyani, D.UmaDevi, “Fraud Detection of Credit
Card Payment System by Genetic Algorithm”,
International Journal of Computer Applications (0975 – 8887)
Volume 161 – No 13, March 2017
9
International Journal of Scientific & Engineering
Research Volume 3, Issue 7, July-2012.
[11] Bidgoli, B. M., Kashy, D., Kortemeyer, G. & Punch, W.
F “Predicting student performance: An Application of
data mining methods with the educational web-based.
(2003). System LON-CAPA”. In Proceedings of
ASEE/IEEE frontiers in education conference. . (2003).
[12] Ekrem Duman, M. Hamdi Ozcelik “Detecting credit card
fraud by genetic algorithm and scatter search”. Elsevier,
Expert Systems with Applications, (2011). 38; (13057–
13063).
[13] Abhinav Srivastava, Amlan Kundu, Shamik Sural and
Arun K. Majumdar, "CreditCard Fraud Detection Using
Hidden Markov Model" IEEE, Transactions On
Dependable And Secure Computing, Vol. 5, No 1. ,
January-March 2008
[14] V. Bhusari, and S. Patil, “Study of Hidden Markov
Model in Credit Card Fraudulent Detection”,
International Journal of Computer Applications (0975 –
8887) Volume 20– No.5, April 2011
[15] V.Bhusari ,S.Patil ," Study of Hidden Markov Model in
Credit Card Fraudulent Detection ",International Journal
of Computer Applications (0975 - 8887) Volume 20-
No.5, April 2011
[16] S. Benson Edwin Raj, A. Annie Portia “Analysis on
Credit Card Fraud Detection Methods”. IEEE-
International Conference on Computer, Communication
and Electrical Technology; (2011). (152-156).
[17] Ray-I Chang, Liang-Bin Lai, Wen-De Su, Jen-Chieh
Wang, Jen-Shiang Kouh “Intrusion Detection by
Backpropagation Neural Networks with Sample-Query
and Attribute-Query”. Research India Publications;
(2006). (6-10).
[18] Raghavendra Patidar, Lokesh Sharma “Credit Card Fraud
Detection Using Neural Network”. International Journal
of Soft Computing and Engineering (IJSCE), (2011).
Volume-1, Issue; (32-38).
IJCA
TM
: www.ijcaonline.org