Summary
Detecting fraud is important but often risks exposing sensitive personal details. The authors found a way to spot suspicious activity by using hidden data patterns instead of direct personal data. Their method filters and then carefully identifies fraud signals while preserving privacy. This approach is very fast and helps organizations quickly flag possible fraud without revealing anyone's private information. It relies on special math techniques that keep data anonymous but still useful for spotting problems.
fraud detectionprivacy preservingtopological anonymizationembeddingunsupervised filteringsupervised snipinglatencyPersonally Identifiable Information (PII)data privacymachine learning
Abstract
Working entirely on topologically anonymized embeddings, we perform fraud detection using iterative rounds of unsupervised filtering followed by supervised sniping. The result is an ultra-low latency privacy--preserving triage that allows institutions to flag suspicious activity without compromising Personally Identifiable Information.