Authentication by Encrypted Negative Password for Credit Card Fraudlent Transaction Detection

DOI : 10.17577/IJERTV10IS020099

Download Full-Text PDF Cite this Publication

Text Only Version

Authentication by Encrypted Negative Password for Credit Card Fraudlent Transaction Detection

Amal Ashok¹, Emy Ann Thomas¹, Nayana V², Chinchu M Pillai², Ajitha Suresh²

U G Students¹,Assistant Professor², Department of Computer Science,

Mount Zion Institute of Science and Technology, Chengannur, Kerala, India

Abstract:- Now on a daily basis the usage of credit cards has dramatically enlarged. As credit card becomes the foremost common mode of payment for each online in addition as regular purchase, cases of fraud related to it are rising. Here, we model the sequence of operations in credit card dealings process employing a Hidden Markov Model (HMM), and it is used for the detection of frauds. An HMM is at the start trained with the conventional behavior of a cardholder. If associate degree incoming credit card dealings are not accepted by the trained HMM with sufficiently high likelihood, it's thought-about to be dishonorable. At the identical time, we to tend to try to make sure that real transaction are not rejected. We have a tendency to gift elaborate experimental results to indicate the effectiveness of our approach and compare it with alternative techniques offered within the literature.


The popularity of online looking is growing day by day. In keeping with AN AC Nielsen study conducted in 2005, tenth part of the worlds population is shopping online. Germany and nice Great Britain has the biggest range of online shoppers, and credit card is the most well-liked mode of payment (59 percent). Regarding 350 million, transactions each year was reportedly applied by Barclay card, the biggest credit card company within the United Kingdom, toward the end of the last century retailers like Walmart usually handle a lot of larger range of credit card transactions as well as online and regular purchases. Credit-card-based purchases are often categorized into 2 types:

-Physical card and

-Virtual card.

In a very physical-card primarily based purchase, the cardholder presents his card physically to a businessperson for creating a payment. To hold out fallacious transactions during this quite purchase, AN offender should steal the credit card. If the cardholder doesn't notice the loss of card, it will cause a considerable loss to the credit card company. Within the second quite purchase, just some vital data a few card (card number, expiration date, secure code) is needed to form the payment. Such purchases square measure ordinarily done on the net or over the telephone. To commit fraud in these varieties of purchases, a fraudster merely must recognize the cardboard details. Most of the time, the real cardholder isn't aware that somebody else has seen or taken his card data. The only way to notice this type of fraud is to investigate the payment patterns on each card and to work out any inconsistency with relevancy the usual payment patterns. Fraud detection supported the analysis of existing

purchase knowledge of cardholder could be a promising way to cut back the rate of booming credit card frauds. Since humans tend to exhibit specific experimental psychology profiles, every cardholder are often pictured by a group of patterns containing data regarding the standard purchase class, the time since the last purchase, the quantity of cash spent, etc. Deviation from such patterns may be a potential threat to the system.


The paper survey, most of the papers searched are used passwords for authentication by using matter password, alpha-numeric password and graphical password this password technique are used however having disadvantage like shoulder surfing brute force attack. Watermarking techniques is additionally used, however not for the password's authentication. Two authentication techniques are supported text and colors projected for PDA during this they generate the session passwords and proof against dictionary attack. Disadvantage of this paper is that each time they generate the session password, and it's tough to recollect new password to the user. Two new authentication schemes certify the user by session passwords that are used just once. Once the session is terminated, the session password is not any longer helpful. For each login method, users input totally different passwords. The session passwords give higher security against dictionary and brute force attacks as password changes for each session. However, during this same drawback is occur that each time user needs to enter password repeat and repeat. It's too hard to recollect password and because it is the session password it's for an actual time only. To remove the disadvantage of textual word removed by graphical word schemes which give the way of creating an excellent deal of easy password, whereas, increasing the number of securities, they are liable to shoulder aquatics. Here text was mix with image and colour to urge the session word and each time user should enter new word as session ends. Two authentication techniques(pair-based authentication pattern and hybrid pattern authentication scheme) for enduring the session passwords. Same disadvantage is here to as antecedent comes. The disadvantages related to the textual passwords like brute-force a lexical attacks and same this drawback management with graphical password that have shoulder surfing and are terribly costly to implement.


System style involves translating data needs and abstract style into technical specification and general flow of process. When the project plan is such, connected information is gathered and how to evolve the system is developed. The method starts with a study of the outputs needed of the system, the inputs given, the data that's to be keep, the way to store the data, the information structured for storage.


Input design is the technique by that valid knowledge is accepted from the user. This a part of the planning needs very careful attention. If the info going into the system is wrong, then the process and output can enlarge these errors. Inaccurate input file is the foremost common reason for errors in processing. Input style consists of the subsequent processes:

-Designing graphical user entry screen is simple to use.

-Designing procedures and functions to valid the data as per business rules. -Designing functions required to store data into a usable type for process.

-Designing the common integrated functions that may be employed by all alternative users when required.


Output design is one of the foremost vital options of the knowledge system. Once the output of fine quality, the users are going to be antipathetic to use the recently designed system and will not use the system. There are many varieties of outputs, all of which may be either extremely helpful or will be vital to the users, counting on the way and degree to that they're used. Outputs from computing system are needed primarily to speak the results of processing to users. They're also wont to give a permanent text of the results for later consultation.


The final and necessary innovate in the system life cycle is the implementation of the new system. The term implementation has totally different meanings starting from the conversion of a basic application to a whole replacement of a computer system. The procedure, however, is nearly a similar. Implementation includes all those activities that take place to convert from previous system to new. The new system could also be altogether new exchange existing system, manual or machine-driven, o it should be a significant modification to an existing system. The method of implementation and duration to be adopted is identified at first. Next, the system is tested properly and at the constant time users are trained within the new procedure. Correct implementation is crucial to produce a reliable system to fulfill organization needs. Successful implementation might not guarantee improvement within the organization using the new system, however, it will prevent improper installation.

Implementation of a new system requires the operating staff installing the software and creating computer files. There are many ways in which this can be achieved. The methods are the following.

  1. Direct change over

  2. Parallel running

  3. Pilot running change

These phase activities do the following:

  1. Development of phase overview

  2. Preparing for implementation

  3. Computer program development

  4. Development phase report and overview There are three types of implementations:

  1. Implementation of a computer system to interchange a manual system. the issues encountered are converting files, coaching users, making correct files and confirmative printouts for integrity.

  2. Implementation of a brand new computer system to exchange an existing one. this can be sometimes a difficult conversion. If not properly planned, there may be several issues. Some massive pc systems have taken as long as a year to convert.

  3. Implementation of a changed application to exchange the prevailing one, using an equivalent computer. This kind of conversion is comparatively straightforward to handle, provided there are not any major changes within the files. each system needs periodic analysis when implementation.


    Today the system is focusing on the analysis of fraudulent transaction by human behavior. At the beginning stage the website is free for all the users. The system has the facility to mind reading technologies.


    In this project, we focus on secure transactions by avoiding fraudulent activities. Here we can make sure that our transaction is to be more accurate and reliable according to the way of transaction. The main idea used here is the concept of negative passwords. Here the passwords should be stored as its encrypted form so the Mallory can difficult to decrypt these passwords.



    1. D. M. Blei, A. Y. Ng, and M. I. Jordan, Latent dirichlet allocation, J. Mach. Learn. Res., vol. 3.

    2. T. Hofmann, Probabilistic latent semantic indexing, in Proc. 22nd Annu. Int. ACM SIGIR Conf. on Res. Develop. Inform. Retrieval.

    3. Y. Gao, Y. Xu, Y. Li, and B. Liu, A two-stage approach for gener-ating topic models, in Advances in Knowledge Discovery and Data Mining.

    4. Y. Gao, Y. Xu, and Y. Li, Pattern-based topic models for informa-tion filtering, in Proc. Int. Conf. Data Min.

    5. S. E. Robertson and I. Soboroff, The TREC 2002 filtering track report,

    6. Y. Cao, J. Xu, T.-Y. Liu, H. Li, Y. Huang, and H.-W. Hon, Adapting ranking svm to document retrieval.

    7. N. Zhong, Y. Li, and S.-T. Wu, Effective pattern discovery for text mining,

    8. X. Yi and J. Allan, A comparative study of utilizing topic models for information retrieval,in Advances in Information Retrieval.

    9. C. Zhai and J. Lafferty, A study of smoothing methods for lan- guage models applied to adhoc information retrieval,.

Leave a Reply