Online Fraud Detection Research Articles

Online fraud detection: ‘In the moment’ digital accountability in a data-sensitive setting

Online fraud detection: ‘In the moment’ digital accountability in a data-sensitive setting

Analysis of the benefits of artificial intelligence and human personality study on online fraud detection

Purpose Technology advancement has changed how banks operate. Modernizing technology has, on the one hand, made it simpler for banks to do their daily business, but it has also increased cyberattacks. The purpose of the study is to to determine the factors that have the most effects on online fraud detection and to evaluate the advantages of AI and human psychology research in preventing online transaction fraud. Artificial intelligence has been used to create new techniques for both detecting and preventing cybercrimes. Fraud has also been facilitated in some organizations via employee participation. Design/methodology/approach The main objective of the research approach is to guide the researcher at every stage to realize the main objectives of the study. This quantitative study used a survey-based methodology. Because it allows for both unbiased analysis of the relationship between components and prediction, a quantitative approach was adopted. The study of the body of literature, the design of research questions and the development of instruments and procedures for data collection, analysis and modeling are all part of the research process. The study evaluated the data using Matlab and a structured model analysis method. For reliability analysis and descriptive statistics, IBM SPSS Statistics was used. Reliability and validity were assessed using the measurement model, and the postulated relationship was investigated using the structural model. Findings There is a risk in scaling at a fast pace, 3D secure is used payer authentication has a maximum mean of 3.830 with SD of 0.7587 and 0.7638, and (CE2). Originality/value This study focused on investigating the benefits of artificial intelligence and human personality study in online transaction fraud and to determine the factors that affect something most strongly on online fraud detection. Artificial intelligence and human personality in the Indian banking industry have been emphasized by the current research. The study revealed the benefits of artificial intelligence and human personality like awareness, subjective norms, faster and more efficient detection and cost-effectiveness significantly impact (accept) online fraud detection in the Indian banking industry. Also, security measures and better prediction do not significantly impact (reject) online fraud detection in the Indian banking industry.

Online banking fraud detection: A comparative study of cases from South Africa and Spain

Background: The banking sector provides online banking to offer their customers convenient and easy access to banking services. As most banking services are nowadays performed online, alarming fraudulent activities occur daily. The online banking frauds are increasingly and commonly being experienced globally and are damaging to both banks and customers.Objectives: The goal of the study was to investigate issues in online fraud detection in the banking sector from South Africa and Spain.Method: The study followed the Design Science Research (DSR) methodology. Data were collected through the qualitative approach particularly using focus groups and semi-structured interviews. The study followed non-probability sampling since it focussed on a specialised area of online fraud model. In total, the population size consisted of 17 participants, which included fraud technical managers, fraud investigator specialists, fraud technical specialists and data scientists.Results: In the South African context, the study established that there is a lack of online fraud experts in the banking sector. The findings reveal that the lack of online fraud expertise may lead to banks having weak detection systems. In the Spanish perspective, the study revealed that there is a lack of law regulation that poses a high risk.Conclusion: Given the lack of online fraud experts in the banking sector, banks are urged to acquire and develop online fraud expertise. Online banking technology is developing expeditiously than traditional transactions; therefore, regulations and policies need to be updated regularly to keep up with rapidly evolving technological changes.Contribution: Considering the significance of global online banking, the study suggested areas that the banking sector may investigate to develop and enhance online fraud detection models to combat online fraudulent activities.

Fraud detection with natural language processing

Automated fraud detection can assist organisations to safeguard user accounts, a task that is very challenging due to the great sparsity of known fraud transactions. Many approaches in the literature focus on credit card fraud and ignore the growing field of online banking. However, there is a lack of publicly available data for both. The lack of publicly available data hinders the progress of the field and limits the investigation of potential solutions. With this work, we: (a) introduce FraudNLP, the first anonymised, publicly available dataset for online fraud detection, (b) benchmark machine and deep learning methods with multiple evaluation measures, (c) argue that online actions do follow rules similar to natural language and hence can be approached successfully by natural language processing methods.

Identifying Fraud Detection Techniques Using Text Analytics Processing

To determine the fraud detection model, to illustrate how the fraud detection model is created, and to start the data model with any classifier, data mining technology is used in the fraud detection process. As e-commerce continues to grow, the associated internet hoax is still a very appealing source of cash for scammers. Because of the severe financial damage that this counterfeit activity does to retailers, online fraud detection is essential. Concerned with scam detection is the need to quickly seize fraudulent actions in addition to containing them. This significance is essential to reducing financial losses. Cybercrimes are a widespread annoyance and have a negative impact on our society in many ways. In every nation, the police enforcement system heavily relies on the results of cybercrime investigations. The connected online hoax is still a highly alluring way for con artists to get money as e-commerce expands. Online fraud detection is crucial due to this counterfeit activity's significant financial harm to merchants. In addition to curbing fraudulent acts, scam identification raises the urgent need to seize them. To minimize financial losses, its relevance is crucial. Cybercrimes are a common irritant that harms our society in a variety of ways. The outcomes of cybercrime investigations substantially influence the police enforcement systems in every country. These methods are essentially utilized for fraud detection in a variety of industries, including health, insurance, online shopping, and more.

MS_HGNN: a hybrid online fraud detection model to alleviate graph-based data imbalance

Online transaction fraud has become increasingly rampant due to the convenience of mobile payment. Fraud detection is critical to ensure the security of online transactions. With the development of graph neural network, researchers have applied it to the field of fraud detection. The existing fraud detection methods will solve the class imbalance by sampling, but they do not fully consider the various imbalances in the heterogeneous graph, and the data imbalance will directly affect the performance of the model. This work proposes a hybrid graph neural network model for online fraud detection to address this issue. The three types of imbalance in online transactions are feature imbalance, category imbalance, and relation imbalance, and they are all addressed in the proposed model. The entities with the feature most closely related to the fraudsters will be determined for the feature imbalance, and samples will be taken for further identification in the subsequent training phase. The hybrid model then uses under-sampling in combination with the long-distance sampling to find nodes with high similarity of features for the category imbalance. Finally, we propose a reward/punishment mechanism based on reinforcement learning for relation imbalance, which uses the threshold created by training as the sampling weight between relations. This paper conducts experiments on the public datasets Amazon and Yelp. The experimental results show that the model proposed is 5.61% higher than the best model in the comparison model on Amazon dataset, and 1.58% higher on Yelp dataset.

NFA: A neural factorization autoencoder based online telephony fraud detection

The proliferation of internet communication channels has increased telecom fraud, causing billions of euros in losses for customers and the industry each year. Fraudsters constantly find new ways to engage in illegal activity on the network. To reduce these losses, a new fraud detection approach is required. Telecom fraud detection involves identifying a small number of fraudulent calls from a vast amount of call traffic. Developing an effective strategy to combat fraud has become challenging. Although much effort has been made to detect fraud, most existing methods are designed for batch processing, not real-time detection. To solve this problem, we propose an online fraud detection model using a Neural Factorization Autoencoder (NFA), which analyzes customer calling patterns to detect fraudulent calls. The model employs Neural Factorization Machines (NFM) and an Autoencoder (AE) to model calling patterns and a memory module to adapt to changing customer behaviour. We evaluate our approach on a large dataset of real-world call detail records and compare it with several state-of-the-art methods. Our results show that our approach outperforms the baselines, with an AUC of 91.06%, a TPR of 91.89%, an FPR of 14.76%, and an F1-score of 95.45%. These results demonstrate the effectiveness of our approach in detecting fraud in real-time and suggest that it can be a valuable tool for preventing fraud in telecommunications networks.

Intelligent mechanical systems and its applications on online fraud detection analysis using pattern recognition K-nearest neighbor algorithm for cloud security applications

Banks and financial institutions use systematic business transaction fraud, individual analytics. Distrust analysis definition is based on a customer history of abnormal behaviors that indicate normal activities and fraud, manipulation or application analysis. The challenge of facing the banking and finance sector is in the form of fraudulent transactions. Combined with the explosive development of data and information, it has urged great advances in technology and communication, global consumer awareness, and capabilities. A comprehensive customer engagement strategy of the banking sector should establish opportunities through dynamic consumer estimation. Security is therefore essential for all banks to implement fraud detection methods to minimize their losses. The regulatory sequence based on different methods of machine learning has evolved to detect various fraudulent transactions. The effective way to detect the fraud detection mechanism is far more essential than the standard classification technology. Pattern Recognition K-Nearest Neighbor [PR-KNN] is developed for such a progressive search and non-parametric technique that evaluates the pseudo-nature to find the best solution to the problem. Regression and classification problems of fraudulent transactions mean the spending is reducing the number of false alarms. The detection systems are placed in bank fraud or fraudulent transactions after the transaction are expected to expectation.

An Analysis of the Supervised Learning Approach for Online Fraud Detection

Illegal online financial transactions are now more sophisticated and global in scope, which costs both parties—customers and businesses. For fraud prevention and detection in the online setting, many different strategies have been proposed. While all of these techniques aim to detect and stop fraudulent online transactions, they differ in terms of their features, advantages, and disadvantages. This study assesses the current fraud detection research in this area to detect the employed algorithms and assessing in accordance with predetermined standards. The systematic quantitative literature review methodology was used to assess the research studies in the subject of online fraud detection. A hierarchical typology is created based on the supervised learning methods in scientific articles and their properties. Therefore, by integrating three selection criteria—accuracy, coverage, and costs—our research presents the best methods for identifying fraud in a novel approach. Index Terms : Detection, Online fraud, Online transaction, Supervised Learning Algorithm.

Mahalanobis distance based accuracy prediction models for Sentinel-2 Image Scene Classification

ABSTRACT Over the years, due to the enrichment of paired-label datasets, supervised machine learning has become a prime component of any problem-solving. Examples include building classifiers for applications such as image/speech recognition, traffic prediction, product recommendation, virtual personal assistant (VPA), online fraud detection and many more. The performance of these developed classifiers is highly dependent upon the training dataset, and subsequently, without human intervention or true labels, the evaluation over unseen observations remains unknown. Using a statistical distance researchers did try to assess the model’s goodness-of-fit and compared multiple independent models. Nonetheless, given a train-test split and different classifiers built over the training set, the question ‘is it possible to find a prediction error using the relation between training and test set?’ remains unsolved. In this article, we propose a generalized statistical distance-based method measuring the prediction uncertainty at a new query point. To be specific, we propose a Mahalanobis distance-based Evidence Function Model to measure the misclassification caused by K-Nearest Neighbours (KNN), Extra Trees (ET), and Convolutional Neural Network (CNN) models when classifying Sentinel-2 image into six scene classes (Water, Shadow, Cirrus, Cloud, Snow, Other). The performance of the proposed method was assessed over two different datasets: (i) the test set, with an overall mean prediction uncertainty detection of 62.99%, 29.80% and 31.51%, leading to a mean micro-F1 performance of 67.89%, 39.30%, and 38.29% for KNN, ET, and CNN, respectively; (ii) a water-body set, with prediction uncertainty detection of 22.27%, 42.08%, and 27.67%, leading to a micro-F1 performance of 34.70%, 58.96%, and 43.32%, respectively.

DEVELOPMENT OF THE AUTOMATED FRAUD DETECTION SYSTEM CONCEPT IN PAYMENT SYSTEMS

The paper presents the demand for the spread of payment systems. This is caused by the development of technology. The open issue of application of payment systems -fraud -is singled out. It is established that there is no effective algorithm that would be the standard for all financial institutions in detecting and preventing fraud. This is due to the fact that approaches to fraud are dynamic and require constant revision of forecasts. Prospects for the development of scientific and practical approaches to prevent fraudulent transactions in payment systems have been identified. It has been researched that machine learning is appropriate in solving the problem of detecting fraud in payment systems. At the same time, the detection of fraud in payment systems is not only to build the algorithmic core, but also to build a reliable automated system, which in real time, under high load, is able to control dataflows and effectively operate the algorithmic core of the system. The paper describes the architecture, principles and operation models, the infrastructure of the automated fraud detection mechanism in payment systems. The expediency of using a cloud web service hasbeen determined. The deployment of the model in the form of automated technology based on the Amazon Web Services platform is substantiated. The basis of the automated online fraud detection system is Amazon Fraud Detector and setting up payment frauddetection workflows in payment systems using a customizable Amazon A2I task type to verify and confirm high-risk forecasts. The paper gives an example of creating an anomaly detection system on Amazon DynamoDB streams using Amazon SageMaker, AWS Glue and AWS Lambda. The automated system takes into account the dynamics of the data set, as theAWS Lambda function also works with many other AWS streaming services. There are three main tasks that the software product solves: prevention and detection of fraud in payment systems, rapid fraud detection (counts in minutes), integration of the software product into the business where payment systems and services are used (for example, payment integration services in financial institutions, online stores, logistics companies, insurance policies, trading platforms, etc.). It is determined that the implementation of an automated system should be considered as a project. The principles of project implementation are offered. It is established that for the rational implementation of the project it is necessary to develop a specific methodology for the implementation of the software product for fraud detection in payment systems of business institutions.

Fast Unsupervised Automobile Insurance Fraud Detection Based on Spectral Ranking of Anomalies

Collecting insurance fraud samples is costly and if performed manually is very time consuming. This issue suggests the usage of unsupervised models for fraud data collection. One of the accurate methods in this regards is Spectral Ranking of Anomalies (SRA) that is shown to work better than other methods for auto-insurance fraud detection, specifically. However, this approach is not scalable to large samples and is not appropriate for online fraud detection. This is while, real-time fraud management systems are necessary to prevent huge losses. In this study, we propose an implementation methodology which makes it possible to apply the SRA to big data senarios. We exploit the power of spectral ranking of anomalies to create an estimated target variable from the unlabeled dataset. We then use two robust models, namely, random forest and deep neural networks to fit a model based on the estimated labeled training set. Next, the incoming live data are fed to the mentioned trained models for predicting the target variable. Simulation results confirm that the proposed approach has higher speed and acceptable false alarm rate compared to existing related methods.

Towards IP geolocation with intermediate routers based on topology discovery

IP geolocation determines geographical location by the IP address of Internet hosts. IP geolocation is widely used by target advertising, online fraud detection, cyber-attacks attribution and so on. It has gained much more attentions in these years since more and more physical devices are connected to cyberspace. Most geolocation methods cannot resolve the geolocation accuracy for those devices with few landmarks around. In this paper, we propose a novel geolocation approach that is based on common routers as secondary landmarks (Common Routers-based Geolocation, CRG). We search plenty of common routers by topology discovery among web server landmarks. We use statistical learning to study localized (delay, hop)-distance correlation and locate these common routers. We locate the accurate positions of common routers and convert them as secondary landmarks to help improve the feasibility of our geolocation system in areas that landmarks are sparsely distributed. We manage to improve the geolocation accuracy and decrease the maximum geolocation error compared to one of the state-of-the-art geolocation methods. At the end of this paper, we discuss the reason of the efficiency of our method and our future research.

Location Accuracy of Commercial IP Address Geolocation Databases

This paper deals with finding the geographical location of Internet nodes remotely with no need to communicate with the nodes located (client-independently). IP geolocation is used in a number of areas, such as content personalisation, on-line fraud prevention and detection, and digital media law enforcement. One of the main concerns when studying the accuracy of client-independent geolocation is the groundtruth dataset. As we show in the related work, the used groundtruth influences the results a lot. We construct an error-free groundtruth dataset consisting of nodes with GPS-precise locations. We also record the country, region, city, and ISP for each groundtruth node. Using the created groundtruth, we study the accuracy of eight IP location databases in a number of scenarios, such as effect of city area and population, effect of ISP assignment, and number of not-returned locations.DOI: http://dx.doi.org/10.5755/j01.itc.46.3.14451

Developing business analytic capabilities for combating e-commerce identity fraud: A study of Trustev’s digital verification solution

Given the significant growth in e-commerce, organizations are seeking novel capabilities and technological innovations to deal simultaneously with the volume of data generated and the need to combat potentially damaging fraudulent activity. Although recent studies identify business analytics (BA) as a potential means of combating fraud, significant inroads into the interrelationships between capabilities and the articulation of a pathway to analytical capability have yet to be made. This study presents an investigation of Trustev, a global provider of digital verification technology, and its development of the profile-based social fingerprinting fraud detection solution. Adopting an interpretive structural modeling technique for data analysis, we construct a framework and reveal a road map for organizations to become analytically capable in online fraud detection. Our study adds to the discourse of the application of BA to combat online fraud.

NuData Security strengthens online fraud detection with behavioural biometrics

NuData Security has strengthened its online fraud detection solution, NuDetect with behavioural biometrics. Through the tracking and monitoring of subconscious behaviours, NuDetect is able to establish a biometric identifier without the need to interrupt a user's experience.

Online Credit Card Fraudulent Detection Using Data Mining

As e-commerce sales continue to grow, the associated online fraud remains an attractive source of revenue for fraudsters. These fraudulent activities impose a considerable financial loss to merchants, making online fraud detection a necessity. The problem of fraud detection is concerned with not only capturing the fraudulent activities, but also capturing them as quickly as possible. This timeliness is crucial to decrease financial losses. In this research, a profiling method has been proposed for credit card fraud detection. The focus is on fraud cases which cannot be detected at the transaction level. In the proposed method the patterns inherent in the time series of aggregated daily amounts spent on an individual credit card account has been extracted. These patterns have been used to shorten the time between when a fraud occurs and when it is finally detected, which resulted in timelier fraud detection, improved detection rate and less financial loss.

About Our Authors

About Our Authors

Bayesian inference – the future of online fraud protection

Fraud is an ever-growing problem for e-commerce businesses, not only because the level of online fraud being committed is rising, but because it is becoming increasingly more complex and sophisticated. As such, if the e-commerce industry is to combat fraud effectively, methods of online fraud detection and prevention also need to grow in sophistication and in erudition. As fraudsters become ever more sophisticated in their methods, so too must the companies that hope to detect and prevent such activities. A major problem for e-commerce businesses today is that fraudulent transactions appear more and more like legitimate ones. The damage to business is not just financial. Other effects can include damaged reputation, loss of trust, lower sales and so on. David Excell, CEO of Featurespace, looks at new technologies based on Bayesian inference that can better protect companies from such increased risks and discusses how they can adapt to new fraud trends.

A novel hybrid artificial immune inspired approach for online break-in fraud detection

Abstract This paper proposes a new hybrid model for online fraud detection of the Video-on-Demand System, which is aimed to improve the current Risk Management Pipeline (RMP) by adding Artificial Immune System (AIS) based fraud detection for logging data. The AIS based model combines two artificial immune system algorithms with behavior based intrusion detection using Classification and Regression trees (CART). Immune inspired algorithms include the improved version of negative selection called Conserved Self Pattern Recognition Algorithm (CSPRA) and a recently established algorithm inspired by Danger Theory (DT) called Dendritic Cells Algorithm (DCA). The hybrid method based on stacking-bagging demonstrates higher detection rate lower false alarm, and handles high dimensional data set better when compared to the results achieved using only CSPRA, DCA, and CART.