A Defense Mechanism for Credit Card Fraud Detection

Abstract

Computer security is one of the key areas where lot of research is being done. Many intrusion detection techniques are proposed to ensure the network security, protect network resources and network infrastructures. Intrusion detection systems (IDS) attempt to detect attacks by gathering network data and analyze the information from various areas to identify the possible intrusions. This paper proposes a defense mechanism combining three approaches such as anomaly, misuse and decision making model to produce better detection accuracy and a decreased false positive rate. The defense mechanism can be built to detect the attacks in credit card system using Hidden Markov approach in the anomaly detection module. The credit card holder’s behaviours are taken as attributes and the anomalous transactions are found by the spending profile of the user. The transactions that are considered to be anomalous or abnormal are then sent to the misuse detection system. Here, the transactions are compared with predefined attack types such as cross site, SQL injection and denial of service attacks and then sent to the decision making model to classify it as known/unknown type of attack. Finally, the decision-making module is used to integrate the detected results and report the types of attacks in credit card system. As abnormal transactions are analyzed carefully in each of the module, the fraud rate is reduced and system is immune to attacks.