Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular Learning

In order to conduct credit card fraud, having only the payment card information of the victim it is possible to fake its identity and buy on e-commerce platforms. This type of fraud is known as Card-Not-Present and shows in the form of chargebacks, projecting billion-dollar losses worldwide in the c...

Full description

Saved in:
Bibliographic Details
Main Authors: Josue Genaro Almaraz-Rivera, Jose Antonio Cantoral-Ceballos, Juan Felipe Botero, Francisco Javier MunOz, Brian David Martinez
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Open Journal of the Computer Society
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11004629/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850129389518323712
author Josue Genaro Almaraz-Rivera
Jose Antonio Cantoral-Ceballos
Juan Felipe Botero
Francisco Javier MunOz
Brian David Martinez
author_facet Josue Genaro Almaraz-Rivera
Jose Antonio Cantoral-Ceballos
Juan Felipe Botero
Francisco Javier MunOz
Brian David Martinez
author_sort Josue Genaro Almaraz-Rivera
collection DOAJ
description In order to conduct credit card fraud, having only the payment card information of the victim it is possible to fake its identity and buy on e-commerce platforms. This type of fraud is known as Card-Not-Present and shows in the form of chargebacks, projecting billion-dollar losses worldwide in the coming years. The IEEE-CIS dataset has emerged as a strong option for creating and validating smart detection systems against this problem. In this work, we propose a solution, Hyphatia, where Self-Supervised Learning is implemented for tabular data based on SubTab. Our model outperforms XGBoost by 2.14% AUROC, detecting 67.44% of the fraud cases in the IEEE-CIS. This pioneering experimentation prioritizes those features that are not obfuscated. Furthermore, beyond providing just classification metrics, we also share time performance and feature importance calculations for explainability. To the best of our knowledge, this is one of the first works in the literature using Self-Supervised Learning for the problem of credit card fraud detection, specifically using the Self-Supervised Tabular Learning approach.
format Article
id doaj-art-96b2833cce0b4181918ea48b7eb214cd
institution OA Journals
issn 2644-1268
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Open Journal of the Computer Society
spelling doaj-art-96b2833cce0b4181918ea48b7eb214cd2025-08-20T02:33:00ZengIEEEIEEE Open Journal of the Computer Society2644-12682025-01-01681282110.1109/OJCS.2025.357060011004629Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular LearningJosue Genaro Almaraz-Rivera0https://orcid.org/0000-0001-8343-4530Jose Antonio Cantoral-Ceballos1https://orcid.org/0000-0001-5597-939XJuan Felipe Botero2https://orcid.org/0000-0002-7072-8924Francisco Javier MunOz3https://orcid.org/0009-0009-5939-2137Brian David Martinez4https://orcid.org/0009-0005-9544-0328Tecnologico de Monterrey, School of Engineering and Sciences, Monterrey, MexicoTecnologico de Monterrey, School of Engineering and Sciences, Monterrey, MexicoUniversidad de Antioquia, Electronics and Telecommunications Engineering Department, Medellin, ColombiaAligo Defensores Informaticos S.A.S., Medellin, ColombiaAligo Defensores Informaticos S.A.S., Medellin, ColombiaIn order to conduct credit card fraud, having only the payment card information of the victim it is possible to fake its identity and buy on e-commerce platforms. This type of fraud is known as Card-Not-Present and shows in the form of chargebacks, projecting billion-dollar losses worldwide in the coming years. The IEEE-CIS dataset has emerged as a strong option for creating and validating smart detection systems against this problem. In this work, we propose a solution, Hyphatia, where Self-Supervised Learning is implemented for tabular data based on SubTab. Our model outperforms XGBoost by 2.14% AUROC, detecting 67.44% of the fraud cases in the IEEE-CIS. This pioneering experimentation prioritizes those features that are not obfuscated. Furthermore, beyond providing just classification metrics, we also share time performance and feature importance calculations for explainability. To the best of our knowledge, this is one of the first works in the literature using Self-Supervised Learning for the problem of credit card fraud detection, specifically using the Self-Supervised Tabular Learning approach.https://ieeexplore.ieee.org/document/11004629/Card-not-present transactionscredit card fraudIEEE-CIS datasetself-supervised learningself-supervised tabular learningSubTab
spellingShingle Josue Genaro Almaraz-Rivera
Jose Antonio Cantoral-Ceballos
Juan Felipe Botero
Francisco Javier MunOz
Brian David Martinez
Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular Learning
IEEE Open Journal of the Computer Society
Card-not-present transactions
credit card fraud
IEEE-CIS dataset
self-supervised learning
self-supervised tabular learning
SubTab
title Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular Learning
title_full Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular Learning
title_fullStr Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular Learning
title_full_unstemmed Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular Learning
title_short Hyphatia: A Card-Not-Present Fraud Detection System Based on Self-Supervised Tabular Learning
title_sort hyphatia a card not present fraud detection system based on self supervised tabular learning
topic Card-not-present transactions
credit card fraud
IEEE-CIS dataset
self-supervised learning
self-supervised tabular learning
SubTab
url https://ieeexplore.ieee.org/document/11004629/
work_keys_str_mv AT josuegenaroalmarazrivera hyphatiaacardnotpresentfrauddetectionsystembasedonselfsupervisedtabularlearning
AT joseantoniocantoralceballos hyphatiaacardnotpresentfrauddetectionsystembasedonselfsupervisedtabularlearning
AT juanfelipebotero hyphatiaacardnotpresentfrauddetectionsystembasedonselfsupervisedtabularlearning
AT franciscojaviermunoz hyphatiaacardnotpresentfrauddetectionsystembasedonselfsupervisedtabularlearning
AT briandavidmartinez hyphatiaacardnotpresentfrauddetectionsystembasedonselfsupervisedtabularlearning