Dynamic Trust Scoring for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Financial institutions and merchants incur significant losses from credit card and e-payment fraud, particularly in e-commerce, where the authenticity of transactions cannot be verified due to the lack of physical payment instruments and unverifiable user identities.
Innovation Solution
A dynamic trust scoring system for mobile devices and networks is implemented, based on device and network parameters such as type, location, transaction history, and behavior, to assess the authenticity of transactions, combining device fingerprinting and IP reputation scores to provide a more accurate assessment of transaction risk.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If dynamic trust scoring system is implemented, then transaction authentication accuracy is improved, but system complexity increases
Solution Approach 1:
The trust scoring system is segmented into distinct components: device trust score (based on device parameters and activities), network trust score (based on IP address reputation), and transaction risk assessment module. Each component operates independently and contributes to the overall authentication decision, making the complex system manageable and maintainable
Solution Approach 2:
The trust scoring system serves multiple functions: it authenticates transactions, assesses risk levels, provides fraud prevention, and enables dynamic updating of device and network reputations. This multi-functional approach consolidates several security functions into a single unified system, reducing overall system complexity despite the enhanced capabilities
2Measurement precision
If multiple trust score parameters are monitored, then fraud detection accuracy is improved, but information processing requirements increase
Solution Approach 1:
The system extracts and monitors only the most critical trust score parameters from the vast amount of available device and network information. Key parameters include device type, registered location, device ID, transaction history, and IP address reputation - filtering out unnecessary data to maintain accurate fraud detection while managing information processing loads
Solution Approach 2:
The system dynamically changes the weight and importance of different parameters based on the specific transaction context and risk level. For example, device location may be more critical for certain transaction types, while transaction history may weigh more for others. This adaptive parameter weighting optimizes fraud detection accuracy while processing manageable information volumes
Data Source
AI summary
A user device is associated with a dynamic trust score that may be updated as needed, where the trust score and the updates are based on various activities and information associated with the mobile device. The trust score is based on both parameters of the device, such as device type, registered device location, device phone number, device ID, the last time the device has been accessed, etc. and activities the device engages in, such as amount of transactions, dollar amount of transactions, amount of denied requests, amount of approved requests, location of requests, etc. Based on a transaction request from the user device, the trust score and a network reputation score is used to determine an overall trust/fraud score associated with the transaction request.