Graphlet-Based Digital Identity Verification for Fraud Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current authentication methods in computer networks face challenges in effectively distinguishing between legitimate and fraudulent users, particularly when authentication credentials are stolen or hacked, and there is a lack of association between credentials and real user identities, leading to risks of fraudulent access.
Innovation Solution
The implementation of a system that uses contextual information, such as device identifiers, user attributes, and geographic location, to create an electronic signature for each access request, correlating this information to enhance user authentication and access control by building a relation graph that connects user identities, devices, and access tokens, thereby improving identity recognition and preventing fraudulent access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional authentication credentials (username/password) are used, then authentication is simple and widely compatible, but the system is vulnerable to credential theft and hacking
Solution Approach 1:
The patent introduces an intermediary authentication mechanism that sits between traditional credentials and access control. This intermediary layer analyzes contextual information (device identifiers, geographic location, user attributes) and generates electronic signatures to verify whether the person presenting credentials is the legitimate owner, thereby enhancing security without completely replacing traditional authentication methods
Solution Approach 2:
The authentication process is segmented into multiple independent verification stages: initial credential verification, contextual information collection, electronic signature generation, and trust score calculation. This segmentation allows the system to maintain compatibility with traditional authentication while adding layered security measures that address credential theft vulnerabilities
2Reliability
If authentication credentials are disconnected from real user identities to allow anonymous access, then user privacy is protected, but the system cannot deny access to fraudulent users based on their real identities
Solution Approach 1:
The system dynamically adjusts the level of identity verification based on risk assessment. For low-risk scenarios, anonymous access is maintained to preserve privacy and usability. For high-risk scenarios or when fraud is detected, the system transitions to verified identity authentication, thereby adapting between anonymity and fraud prevention based on contextual conditions
Solution Approach 2:
The patent changes the parameter of identity verification from a static binary state (anonymous or verified) to a dynamic spectrum based on trust scores and reputation analysis. The system can adjust the degree of identity association required based on the calculated risk level, allowing flexible balancing between privacy protection and fraud prevention
3Measurement precision
If contextual information is collected and analyzed to create electronic signatures, then user identification accuracy is improved, but data processing complexity and computational resources increase
Solution Approach 1:
The system performs preliminary collection and hashing of contextual information (device identifiers, user attributes, geographic location) during user registration or initial access. This preliminary action creates pre-computed electronic signatures that can be quickly compared during authentication, reducing the computational burden during actual access control decisions
Solution Approach 2:
Instead of storing and processing large volumes of raw contextual data, the system creates simplified copies in the form of electronic signatures and trust scores. These condensed representations capture the essential verification information while significantly reducing data processing complexity and storage requirements
4Reliability
If reputation scores and trust scores are calculated for access tokens and user identities, then fraudulent access is prevented, but authentication processing time increases
Solution Approach 1:
Reputation scores and trust scores are updated periodically based on user behavior patterns and access history, rather than being recalculated in real-time for every authentication attempt. This periodic update mechanism maintains current security assessments while minimizing processing time during actual access control operations
Solution Approach 2:
Trust scores and reputation metrics are pre-calculated and stored based on historical user behavior and access patterns. During authentication, the system performs quick lookups of these pre-computed scores rather than conducting full analysis, thereby maintaining high security while reducing authentication processing time
Data Source
AI summary
A graphlet is extracted from a relation graph for digital identity verification. The relation graph contains data element nodes collected about access activities, such as user access context, payment instrument, address, device information, etc. The graph can be examined to identify node groups in demarcation areas. Nodes outside of the demarcation areas represent data elements of a user; and the data elements in the demarcation areas may or may not be of the user in general. The system determines a matching score between the nodes outside of the demarcation areas and the nodes in a demarcation area. The nodes in the demarcation area are attributed to the user if the matching score is above a threshold. The nodes attributed to the user are combined as the graphlet showing the data elements of the user and used in user authentication and/or access control via electronic signature.