Activated Neural Pathways in Knowledge Graphs for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to efficiently analyze and classify user account activities, particularly in unstructured data formats, for fraud detection and security policy violations, due to computational complexity and the need for real-time processing.
Innovation Solution
The use of graph-structured data models, specifically knowledge graphs, to generate interlinked descriptions of entities and calculate multi-node similarity and sentiment scores, allowing for the identification of 'neural pathways' that indicate potential fraud, thereby reducing data processing complexity and enabling near real-time adjudication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional machine learning methods are used to analyze user account activities, then comprehensive data analysis can be performed, but computational complexity increases and real-time processing becomes difficult
Solution Approach 1:
The patent segments the knowledge graph analysis by identifying and focusing only on 'activated neural pathways' - specific subsets of nodes and edges that are relevant to the query. This segmentation reduces the computational complexity from analyzing the entire knowledge graph to analyzing only the relevant pathways, while maintaining comprehensive analysis of important data relationships.
Solution Approach 2:
The patent extracts and focuses on specific critical elements (activated neural pathways) from the larger knowledge graph structure. By identifying pathways where node activations exceed threshold values, the system extracts only the most relevant data relationships for fraud detection, eliminating unnecessary computational operations on irrelevant data.
2Measurement precision
If traditional data processing methods are used, then thorough analysis can be conducted, but service delays increase
Solution Approach 1:
The patent performs preliminary action by pre-computing and storing node activation values and pathway information in the knowledge graph structure. When a classification query is received, the system can quickly retrieve and analyze pre-computed pathway data rather than performing full data processing from scratch, enabling near real-time responses while maintaining accurate classification.
Solution Approach 2:
The patent applies partial action by analyzing only the necessary subset of data (activated neural pathways with threshold-exceeding activations) rather than processing the entire knowledge graph. This selective analysis reduces processing time significantly while maintaining classification accuracy by focusing computational resources on the most relevant data elements.
3Measurement precision
If comprehensive knowledge graph analysis is performed, then fraud detection accuracy improves, but unnecessary operations increase processing time
Solution Approach 1:
The patent applies local quality by differentiating between activated and non-activated neural pathways based on threshold comparisons of node activation values. The system applies different processing treatments: detailed analysis for activated pathways (where node activations exceed thresholds) and minimal or no processing for non-activated pathways, thereby improving efficiency without sacrificing detection accuracy in critical areas.
Solution Approach 2:
The patent performs partial action by conducting thorough analysis only on activated neural pathways that meet activation thresholds, while performing minimal or no analysis on non-activated pathways. This selective approach maintains high fraud detection accuracy for relevant cases while significantly improving overall data processing efficiency by avoiding unnecessary operations on irrelevant data.
Data Source
AI summary
Systems and methods for determining activated neural pathways in knowledge graphs are disclosed. In an embodiment, a computer system may retrieve data associated with a user account and generate a knowledge graph centered around an account under a current adjudication. The computer system may determine paths based on the knowledge graph using similarity and sentiment scores. By reducing the amount of data in the knowledge graph to specific data associated with paths of interest, unnecessary computer operations may be avoided, which increases processing times related to the knowledge graph and allows near real-time outputs based on the knowledge graph to be utilized. The computer system may rank the paths based on their similarity scores and sentiments scores to determine which paths are considered activated neural pathways in the knowledge graph. The computer system may base a determination of the current adjudication for the account on the activated neural pathways.