Mobile Device Fraud Detection via Behavioral Pattern Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods fail to effectively identify and prevent fraudulent use of mobile devices, despite increased security measures, as they lack a comprehensive system to assess abnormal behavior and initiate corrective actions.
Innovation Solution
A system and method that utilize group associations, proximity identification, and location tracking to determine normal behavior patterns of mobile devices, flagging deviations as potential fraudulent activity by comparing historic data with current interactions and locations, and employing confidence levels and thresholds to monitor and assess suspicious behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple security measures (embedded agents, passwords, encryption) are implemented to improve transaction security, then the reliability of transactions is improved, but the device complexity increases
Solution Approach 1:
The patent introduces a server-based fraud detection system that acts as an intermediary between mobile devices and transaction processors. This external system analyzes behavioral data from multiple devices to identify fraudulent patterns, thereby improving transaction security without increasing the complexity of individual mobile devices or their embedded security measures.
2Reliability
If location-based restrictions are implemented to prevent fraudulent operations, then transaction security is improved, but the adaptability of the system to different user behaviors deteriorates
Solution Approach 1:
The patent implements dynamic fraud detection that adapts to changing user behaviors and patterns. Instead of static location-based restrictions, the system continuously learns from historical data, identifies abnormal behavioral patterns, and adjusts its detection criteria accordingly. This allows the system to maintain high security while adapting to legitimate changes in user behavior, such as traveling to new locations or changing daily routines.
3Measurement precision
If comprehensive behavioral monitoring is implemented to identify fraudulent use, then the measurement precision of fraud detection is improved, but the loss of user information privacy increases
Solution Approach 1:
The patent utilizes universally collected mobile device data (location, group associations, proximity information) that is already being gathered for legitimate communication purposes. By analyzing this existing multi-functional data through behavioral pattern recognition, the system achieves high fraud detection accuracy without requiring additional intrusive data collection, thereby minimizing privacy intrusion while maintaining detection precision.
Data Source
AI summary
Method of determining fraudulent use based on behavioral abnormality starts with processor receiving first location data and first proximity information from first mobile device. First proximity information includes identification of mobile devices within proximity sensitivity radius of first mobile device. Processor determines whether first location data and first proximity information are included in historical location data and historical proximity information, respectively, associated with first mobile device. When first location data and first proximity information is not included, processor determines whether subsequent location data and subsequent proximity information received from first mobile device over predetermined period of time is included. Processor signals to monitor fraudulent use of first mobile device when subsequent location data and subsequent proximity information received from first mobile device over predetermined period of time is not included in historical location data and historical proximity information, respectively, associated with first mobile device. Other embodiments are described.