Analysis of Fraudulent Transaction Detection Techniques based on Customer Behavioural Patterns

Rapid Growth of internet and E-Commerce has opened a new Marketing Phenomenon i.e Online Shopping. However, the problem of same being illegally used by Intruders for Fraudulent activities has become increasingly prominent, and the security of user account is not guaranteed. On the other side, the e-commerce website keeps all the data record related to transaction so there is tremendous amount of data identifying human behaviours. This paper is based on study of different methodology used for identifying the fraudulent transaction with different attributes of users based on their historical transaction. Each historical transaction consists of various transaction attributes which can be used to detect whether the transaction is fraud or genuine. In each of the methodology Initial step is to create a behavioural profiling model by assigning score and then classify the transaction as fraudulent or nonfraudulent transaction.