Mouse Movement Biometrics for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional fraud detection systems are vulnerable to sophisticated man-in-the-browser attacks, as malware can mimic customer profiles by gathering information on typical transaction patterns, making it difficult to distinguish between legitimate and fraudulent transactions.
Innovation Solution
Incorporating mouse movement data into the user profile by training a fraud detection system to differentiate between legitimate and fraudulent users using mouse movement datasets, which are clustered based on distance and cluster metrics, providing a robust framework for fraud detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional fraud detection systems use transaction attributes (time, IP address, geolocation) for training, then the system can identify basic fraud patterns, but the system becomes vulnerable to MITB attacks where malware can mimic customer profiles
Solution Approach 1:
The patent extracts the biometric element (mouse movement patterns) from the traditional fraud detection system that relies solely on transaction attributes. By separating the authentication mechanism from transaction data and using mouse movement biometrics instead, the system removes the vulnerability point where malware could mimic customer profiles based on transaction patterns alone
Solution Approach 2:
The patent replaces the mechanical/system-based fraud detection (analyzing transaction attributes like IP addresses and timestamps) with a biometric-based system that analyzes mouse movement patterns. This substitution makes detection more reliable because biometric data is much harder for malware to replicate compared to structured transaction attributes
2Reliability
If the fraud detection system collects and analyzes mouse movement data, then the system can differentiate between legitimate and fraudulent users, but the system complexity increases
Solution Approach 1:
The patent makes the mouse input device serve multiple functions: it is both the interface for normal user interaction with the system and the source of biometric data for fraud detection. The same mouse movements used for navigation and data entry are simultaneously analyzed for authentication purposes, eliminating the need for separate authentication hardware or interfaces
Solution Approach 2:
The system uses the user's own natural mouse movements as the authentication mechanism. The biometric data is collected passively during normal interaction without requiring the user to perform separate authentication actions, enroll biometrics, or change their natural behavior patterns
Data Source
AI summary
An improved technique trains a fraud detection system to use mouse movement data as part of a user profile. Along these lines, a training apparatus receives sets of mouse movement datasets generated by a legitimate user and/or a fraudulent user. The training apparatus assigns each mouse movement dataset to a cluster according to one of several combinations of representations, distance metrics, and cluster metrics. By correlating the clusters with the origins of the mouse movement datasets (legitimate or fraudulent user), the training apparatus constructs a robust framework for detecting fraud at least partially based on mouse movement data.