The TACC research team from National Chengchi University has published the paper "Credit Card Fraud Detection via Intelligent Sampling and Self-supervised Learning" in ACM Transactions on Intelligent Systems and Technology. In the post-pandemic era, the rapid growth of online shopping and digital payments has made credit card fraud detection increasingly crucial. The team addresses the key challenges in current fraud detection systems, including inadequate transaction representation, noise labels, and data imbalance. The team designed a fraud detection method that accurately extracts spatial and temporal representative features, utilizing intelligent sampling strategies and self-supervised learning techniques to enhance performance. When applied to real-world datasets, the results demonstrated a higher F1 score compared to the most commonly used online fraud detection methods, making a significant contribution to risk management for financial institutions.

 (1) This work has been accepted by ACM Transactions on Intelligent Systems and Technology (TIST) 

(2) DOI Bookmark: https://doi.org/10.1145/3641283