Affle 3i receives US patent for its fraud detection engine
The patent's subject area is ?Method and System for Hardware and Software Based User Identification for Advertisement Fraud Detection?. It introduces a smart framework for detecting digital advertisement fraud and online commerce fraud, by combining plurality of hardware and software level data to distinguish genuine human actions from bot-initiated or manipulated activity. Unlike conventional approaches that rely solely on software identifiers, this system also analyses multi-layered data in real-time through various hardware touchpoints like device sensors, biometrics and connectivity signals. The fraud detection engine builds unique device/user behavioural profiles and calculates probabilistic scores by correlating data with pre-defined baselines, cross-device mapping and digital fingerprinting techniques.
The system's real-time design employs supervised & unsupervised machine learning and intelligent algorithms to adapt to the continuously evolving fraud patterns. It also analyses data from third-party connected devices (such as IoT, smart wearables, connected TVs, etc.) which are assigned unique identities and communicate via wireless protocols. By correlating this data with sensor-based indicators such as gyroscope, accelerometer data, touch interactions and other user device signals, the system strengthens cross-device fraud detection and enables blocking of fraudulent publishers/devices in real-time. This significantly reduces cost leakages and optimises campaign efficiencies at scale across a plurality of connected devices.
This marks the 16th patent granted to Affle.
The inventors of this patent include Anuj Khanna Sohum, Charles Yong and Anurag Singh.
Powered by Capital Market - Live News