Fraud Detection in Mobile Payment Systems using an XGBoost-based Framework

2023 | Information Systems Frontiers | Citations: 0

Authors: Hajek, Petr; Abedin, Mohammad Zoynul; Sivarajah, Uthayasankar

Abstract: Mobile payment systems are becoming more popular due to the increase in the numb ... Expand

Abstract: Mobile payment systems are becoming more popular due to the increase in the number of smartphones, which, in turn, attracts the interest of fraudsters. Extant research has therefore developed various fraud detection methods using supervised machine learning. However, sufficient labeled data are rarely available and their detection performance is negatively affected by the extreme class imbalance in financial fraud data. The purpose of this study is to propose an XGBoost-based fraud detection framework while considering the financial consequences of fraud detection systems. The framework was empirically validated on a large dataset of more than 6 million mobile transactions. To demonstrate the effectiveness of the proposed framework, we conducted a comparative evaluation of existing machine learning methods designed for modeling imbalanced data and outlier detection. The results suggest that in terms of standard classification measures, the proposed semi-supervised ensemble model integrating multiple unsupervised outlier detection algorithms and an XGBoost classifier achieves the best results, while the highest cost savings can be achieved by combining random under-sampling and XGBoost methods. This study has therefore financial implications for organizations to make appropriate decisions regarding the implementation of effective fraud detection systems. Collapse

Topics: fraud detection mobile system mobile payment electronic payment Python

Methods: XGBoost machine learning anomaly detection supervised learning experiment

CATCHM: A novel network-based credit card fraud detection method using node representation learning

2023 | Decision Support Systems | Citations: 0

Authors: Van Belle, Rafaël; Baesens, Bart; De Weerdt, Jochen

Abstract: Advanced fraud detection systems leverage the digital traces from (credit-card) ... Expand

Abstract: Advanced fraud detection systems leverage the digital traces from (credit-card) transactions to detect fraudulent activity in future transactions. Recent research in fraud detection has focused primarily on data analytics combined with manual feature engineering, which is tedious, expensive and requires considerable domain expertise. Furthermore, transactions are often examined in isolation, disregarding the interconnection that exists between them. Collapse

Topics: fraud detection credit card fraud privacy business process management electronic mail

Methods: computational algorithm experiment machine learning feature engineering literature study

Is the Repeal of Net Neutrality a Necessary Evil? An Empirical Analysis of Net Neutrality and Cybercrime

2022 | International Conference on Information Systems | Citations: 0

Authors: Kim, Doehun; Park, Jiyong

Abstract: While net neutrality guarantees equal access to the Internet and online content ... Expand

Abstract: While net neutrality guarantees equal access to the Internet and online content, it serves as a limiting factor in identifying and tracking criminal activities in cyberspace by ensuring that data packet is transmitted with equal priority, irrespective of its source and content. Exploiting a natural experiment in which net neutrality policies were officially repealed in 2018 in the United States, this study examines the impact of net neutrality on the occurrence of cybercrime. Our findings suggest that the repeal of net neutrality is negatively associated with the occurrence of malicious code and content in an attempt to compromise computer systems (e.g., malware and ransomware). In contrast, we do not find any significant relationship with cybercrime victimization, and cybercrime that may subsequently occur in compromised systems (e.g., data breaches and denial-of-service attacks). This study provides novel insights into the role of net neutrality and open Internet toward the preventive cybersecurity paradigm. Collapse

Topics: net neutrality computer crime internet technology IT security criminality

Methods: experimental group natural experiment robustness check longitudinal research qualitative interview

Theories: economic theory

Improving Imbalanced Machine Learning with Neighborhood-Informed Synthetic Sample Placement

2022 | Journal of Management Information Systems | Citations: 0

Authors: Nasir, Murtaza; Dag, Ali; Simsek, Serhat; Ivanov, Anton; Oztekin, Asil

Abstract: Machine learning is widely used in information systems design. Yet, training alg ... Expand

Abstract: Machine learning is widely used in information systems design. Yet, training algorithms on imbalanced datasets may severely affect performance on unseen data. For example, in some cases in healthcare, fintech, or cybersecurity contexts, certain subclasses are difficult to learn because they are underrepresented in training data. Our study offers a flexible and efficient solution based on a new synthetic average neighborhood sampling algorithm (SANSA), which, in contrast to other solutions, introduces a novel "placement" parameter that can be tuned to adapt to each dataset's unique manifestation of the imbalance. This package can be downloaded for R1. We tested SANSA against seven existing sampling methods used in conjunction with the four most frequently used machine learning models trained on 14 benchmark datasets. Our results provide suggestive evidence that SANSA offers a feasible solution to the imbalance problem for most datasets. Our findings provide practical recommendations for how SANSA can be effectively implemented while reducing the complexity level of an imbalanced learning pipeline. Collapse

Topics: marketing management customer service online review logistics management credit card fraud

Methods: computational algorithm machine learning support vector machine artificial neural network descriptive statistic

False Positives in Credit Card Fraud Detection: Measurement and Mitigation

2022 | HICSS | Citations: 0

Authors: Wallny, Florian

Abstract: Credit Card Fraud Detection is a classification problem where different types of ... Expand

Abstract: Credit Card Fraud Detection is a classification problem where different types of classification errors cause different costs. Previous works quantified the financial impact of data-driven fraud detection classifiers using a cost-matrix based evaluation approach, however, none of them considered the significant financial impact of false positives. Analysts reported that fraud prediction in e-commerce still has to deal with false positive rates of 30-70%, and many cardholders reduce card usage after being wrongly declined. In our paper, we propose a new method for assessing the cost of false positives and evaluate several state-of-the-art fraud detection classifiers using this method. Further, we investigate the effectiveness of ensemble learning as previous work supposed that a combination of diverse, individual classifiers can improve performance. Our results show that cost-based evaluation yields valuable insights for practitioners and that our ensemble learning strategy indeed cuts fraud cost by almost 30%. Collapse

Topics: fraud detection credit card fraud data warehouse

Methods: machine learning random forest classification XGBoost feature engineering ensemble learning

Assessing Privacy and Security of Information Systems from Audit Data

2022 | Information Systems Frontiers | Citations: 0

Authors: Westland, J. Christopher

Abstract: We investigated publicly reported security breaches of internal controls in corp ... Expand

Abstract: We investigated publicly reported security breaches of internal controls in corporate information systems to determine whether U.S. Securities and Exchange Commission (SEC) data are information bearing with respect to breaches of security and privacy. The issue has grown in importance as information systems breaches have steadily grown costlier and more frequent. Our analysis supports a high predictability for credit card breaches, portable device related breaches and breaches conducted by firm insiders. Our study also found evidence that employees are subverting particularly strict internal controls by using portable devices that can be carried outside the physical boundaries of the firm. In general, auditing and corporate data filed with the SEC was non-informative with regard to breaches involving unintended disclosures, physical losses, hacking and malware and workplace computers. Scope and fees associated with auditing are significant factors in predicting security breaches, whereas assessments of internal controls effectiveness was shown to be less significant for prediction. Collapse

Topics: Sarbanes–Oxley Act IT security threat credit card fraud privacy IT security

Methods: machine learning random forest classification gradient boosting general multivariate regression model XGBoost

Digital and Physical Mechanisms of Financial Frauds and Identity Theft

2021 | International Conference on Information Systems | Citations: 0

Authors: Qian, Junjie (Henry); Tanriverdi, Hüseyin

Abstract: We extend the routine activity theory (RAT) by developing the novel concept of ... Expand

Abstract: We extend the routine activity theory (RAT) by developing the novel concept of proximity to identity. We theorize that individual's personal information falls into four identity dimensions, including symbolic representation, personal identity, role identity and social identity. Building on this theoretical framework, we distinguish the level of proximity to identity created by three channels that can leak personal information, including individual online shopping activity, individual physical theft, and organizational data breach. Using a representative sample of 23,376 U.S. citizens, we find that individual online shopping activity and individual physical theft increase the probability of victimization that require low to medium proximity to identity, including credit and debit card fraud, other existing account fraud and new account fraud. We also find that organizational data breach increases the probability of all types of identity frauds and identity theft victimization. Collapse

Topics: identity theft electronic commerce data breach criminality credit card fraud

Methods: survey longitudinal research theory development data transformation principal component analysis

Theories: activity theory

Feature construction for fraudulent credit card cash-out detection

2019 | Decision Support Systems | Citations: 0

Authors: Wu, Yue; Xu, Yunjie; Li, Jiaoyang

Abstract: As a type of credit card fraud behavior, fraudulent cash-out causes banks and co ... Expand

Abstract: As a type of credit card fraud behavior, fraudulent cash-out causes banks and companies to incur huge losses. Furthermore, such organizations lack effective methods to detect fraudulent cash-out. To detect fraudulent cash-out accurately, we construct four feature sets based on domain knowledge from industry experts, tips from fraudsters, related media reports, and the relevant literature. In addition, we construct features using a functional data analysis algorithm to capture the time-dependent behavioral patterns of cardholders. We also construct a benchmark feature set based on the traditional approach of Whitrow's strategy. We compare these feature sets using a real data set comprising real transactions of 25,000 credit cards with three machine learning methods: eXtreme gradient boosting, random forest, and support vector machine. The results reveal that our proposed features, which consider both snapshot and dynamic behavioral patterns of cardholders, achieve considerably superior performance to that of Whitrow's strategy. The precisions at the top 5%, 10%, 15%, and 20% levels are improved by magnitudes of 0.049, 0.081, 0.053, and 0.046, respectively. Collapse

Topics: credit card fraud fraud detection website online community security policy

Methods: computational algorithm random forest classification support vector machine experiment classification method

Classifying RFID attacks and defenses

2018 | Americas Conference on Information Systems | Citations: 0

Authors: Mitrokotsa, Aikaterini; Rieback, Melanie R.; Tanenbaum, Andrew S.

Abstract: Fraud related studies cover a wide range of approaches for detection and identif ... Expand

Abstract: Fraud related studies cover a wide range of approaches for detection and identifying or mitigating risks associated with it. This study extends the existing body of research and our understanding on fraud with a specific focus on transit media. A large public transit transaction dataset was extracted from production systems of a large U.S Public Transit Authority system. We developed various machine learning models using techniques such as random forests, support vector machines, and artificial neural networks to effectively classify transit fare media fraud into binary groups and compare their relative model performance. We found that random forests and neural networks outperformed support vector machines for transit fraud detection. More data and parameter optimization are needed for improved results. Collapse

Topics: fraud detection radio frequency identification credit card fraud manufacturing information system

Methods: machine learning random forest classification artificial neural network support vector machine supervised learning

Computer assisted frauds: An examination of offender and offense characteristics in relation to arrests

2017 | Information Systems Frontiers | Citations: 7

Authors: Liao, Ruochen; Balasinorwala, Shenaz; Raghav Rao, H.

Abstract: Previous studies on fighting computer-assisted frauds have attempted to assist l ... Expand

Abstract: Previous studies on fighting computer-assisted frauds have attempted to assist law enforcement agencies (LEAs) to better understand important aspects of motivation, opportunity and deterrence. However, there have been few empirical studies on the profiles of convicted offenders, post detection. This paper examines characteristics of frauds and their associated respective law enforcement response with particular emphasis on frauds facilitated by information technology. The findings show how the prosecution and conviction of the offenders differ among commonly-seen types of computer assisted frauds, and bring new evidence to the common association of gender and crime, severity and punishment, etc. The findings may help LEAs and legislative bodies to evaluate their current practices from the point of restorative social justice. Collapse

Topics: criminality internet technology logistics management accounting identity theft

Methods: descriptive statistic logistic regression regression analysis method statistical hypothesis test