Contact Graph Scoring for Risk Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine learning risk vetting platforms for electronic services often fail to accurately assess users without device history, transaction history, or credit history, as they do not consider the risk behaviors of the user's contacts, leading to incorrect categorization and potential denial of services.
Innovation Solution
A contact graph scoring system that uses a machine learning approach to generate a user score based on the behaviors of their contacts, training a contact graph model using labeled contact graphs to provide a more accurate risk assessment and compliance evaluation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional machine learning risk vetting platforms are used, then users with device history, transaction history, or credit history can be assessed, but users without such history cannot be accurately assessed and may be incorrectly categorized as high risk
Solution Approach 1:
The patent introduces a new dimension for risk assessment by analyzing contact graphs instead of relying solely on traditional user history data. The system constructs a graph where users are nodes and their contacts are edges, enabling risk assessment through social network relationships rather than individual user behavior history.
Solution Approach 2:
The patent uses contact information as an intermediary to assess user risk. Instead of directly analyzing user behavior history, the system examines the risk profiles of users' contacts to infer the target user's risk level, thereby bridging the gap for users without traditional history.
2Adaptability or versatility
If contact graph analysis is implemented, then users without traditional history can be accurately assessed, but the system complexity increases
Solution Approach 1:
The contact graph scoring system serves multiple functions: it assesses user risk, evaluates compliance, and personalizes services. By building a universal contact graph representation, the system can apply the same underlying structure to various assessment purposes, reducing overall system complexity despite the added capability.
Solution Approach 2:
The system transforms contact list data into a structured contact graph representation with specific parameters (contact type, relationship strength, risk level). This parameterization allows complex social network analysis to be reduced to standardized computational operations, managing system complexity through abstraction.
3Measurement precision
If contact graph model training is performed, then risk assessment accuracy improves, but the time required for model training and scoring increases
Solution Approach 1:
The system performs preliminary model training using historical contact graph data and labeled risk information. Once trained, the model can quickly assess new users without requiring time-consuming real-time analysis, as the complex pattern recognition is pre-computed during training.
Solution Approach 2:
The system creates a trained contact graph model that captures risk patterns from historical data. This trained model acts as a copy of the complex analysis logic, enabling fast inference on new users without reperforming the comprehensive training analysis each time.
Data Source
AI summary
Machine learning techniques are disclosed that allow device contact list information to be leveraged in building better models that provide more accurate assessment of user transaction risks. A computing device may receive a contact list that includes a first set of user device identifiers and generates a contact graph for that user by associating the user device identifier with the first set as first-degree contacts. The computing device may then determine that a portion of the first set of the user device identifiers are stored in a contact database (e.g. on a server) and generate a user score based on user information associated with the first set, the contact graph, and a contact graph model. The computing device may provide the user score to a transaction assessment service as in input for the transaction assessment service deciding whether particular electronic transactions are approved for processing by an electronic service provider.