Abstract
Some systems for detecting fraudulent user-generated content (UGC) can be siloed and reactive, analyzing individual pieces of content or account metrics in isolation, which may leave them susceptible to coordinated campaigns spanning multiple services. A system and method are described that can construct a unified cross-service behavioral graph to address such challenges. This graph can ingest and correlate metadata from user activities across disparate services, for example, productivity, payment, and communication platforms, to identify precursor signals that may be indicative of fraudulent intent. By analyzing patterns such as co-reviewing, temporal pulsing, and suspicious cross-service coordination (e.g., template creation), the system can identify coordinated inauthentic behavior. This approach can facilitate a shift from reactive analysis toward pre-emptive detection, potentially improving the identification of large-scale UGC fraud, in some cases before malicious content is published.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Kumar S R, Mithun, "Correlating Cross-Service Behavioral Signals to Identify Coordinated Inauthentic Activity", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/8499