Article Preview
Top1. Introduction
The credit card is the most important payment worldwide due to the expeditious development in information technology. The credit card is utilized in the form of tender for obtaining services or goods, and it is one of the familiar types of fraud. Usually, it is created with the theft of real account number or the credit card, but identifying the fraud is very complex (Darwish, 2020). Normally, fraud is shopping with a card without the knowledge of the owner. Thus, billion-dollar losses happened due to the insecure usage of credit cards (Taha & Malebary, 2020; Zhang et al., 2019). The accurate estimation regarding the losses is quite challenging to predict because credit card firms are normally disinclined to declare the information. However, credit card utilization with a lack of strong security originates the billion-dollar economic loss (Carneiro et al., 2017; Taha & Malebary, 2020). A huge amount is gained with the threat in a short period, so a credit card is measured as the main intention of fraud (Darwish, 2020). The fraud can be employed in various mechanisms, online, offline, counterfeit, and application fraud. Normally, two kinds of fraud happen behavior fraud (Bolton & Hand, 2002; Taha & Malebary, 2020) and application fraud (Phua et al., 2019). The wrong credit card is applied with the wrong information, and the issuer provides the credit card. Therefore, it is termed application fraud. Behavior fraud is the theft in credit card transactions, which entails fake activities (Taha & Malebary, 2020).
In addition, there are two techniques used to conflict the fraud, namely fraud identification and its prevention. Fraud preservation by filtering high-risk transactions and it is an initial line of security. In addition, there are several authorization systems, like cardholder's address, identification number, signatures, expiry data, and signatures are utilized for credit card fraud preservation (Abdallah et al., 2016; Yang et al., 2019). Normally, detection of fraud is separated based on misuse detection and anomaly detection (Zheng et al., 2013; Zhu et al., 2020). The anomaly detection utilizes the formal transaction to identify the misuser, whereas the misuse detection employs the labeled transaction for the detection of misuser (Seyedhossein & Hashemi, 2010; Pumsirirat & Yan, 2018). The constantly increasing cash-less economy leads to business movement to electronic monetary transactions, which is precisely used to identify fraud (Yang et al., 2019). Credit card transactions are commonly taking place by the development of global communication and modern computing technologies. The manually designed rules, such as subjective, in-efficient, and time-consuming, are utilized by the fake transaction detection. When a technique identifies the fraud, the total loss is decreased radically (Kim et al., 2019). The manual verification of all transactions is difficult for the provider because of the enormous number of transactions (Zhang et al., 2019). A good fraud detection method can detect the fraud transactions correctly, and the detections are created in real time transactions. Anomaly detection technologies are normally used to recognize new credit card frauds (Pumsirirat & Yan, 2018).