New CATeye framework combats evolving e-commerce voucher fraud

Researchers have developed a new framework called CATeye to combat evolving voucher abuse in e-commerce. This method addresses the challenge of coupled attribute-topology shifts, where changes in attribute proximity can alter the graph's topology, amplifying detection errors in graph neural networks. CATeye utilizes an Attribute Invariance Selector to filter irrelevant attributes and an Edge Invariance Selector to isolate invariant subgraphs, thereby constructing multiple views for robust domain-invariant representation learning. Experiments on a proprietary dataset from Lazada and a public benchmark demonstrated CATeye's superiority, achieving up to an 8.61% improvement in average F1 score over existing methods. AI

IMPACT This research offers a novel approach to detecting evolving fraud patterns in e-commerce, potentially improving the accuracy and robustness of detection systems.

RANK_REASON Academic paper detailing a new framework and experimental results. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• Attribute Invariance Selector
• CATeye
• Edge Invariance Selector
• graph neural network
• Lazada Group

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →