Graph Path Cache for Real-Time ML Data Consumption
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional implementations of graph databases with machine learning models require substantial computational resources to meet service level agreements (SLAs) for real-time data consumption, especially with dense vertices, and often cannot scale horizontally to support both real-time and batch processing due to computational constraints.
Innovation Solution
A method utilizing a graph path cache for precomputing graph paths, which allows for real-time data consumption by machine learning models by performing deep link analysis and executing models using precomputed graph attributes, enabling efficient processing of high-density vertices and facilitating identification of synthetic identity fraud.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional graph database implementations are used with machine learning models, then real-time data consumption can be achieved, but substantial computational resources are required and horizontal scaling is limited
Solution Approach 1:
The patent precomputes graph paths between vertices and stores them in a cache before real-time processing. This preliminary computation eliminates the need for expensive real-time path calculations, significantly reducing computational resources required during actual machine learning model execution while maintaining real-time data consumption capability
Solution Approach 2:
The patent creates cached copies of precomputed graph path data that can be rapidly accessed during real-time processing. Instead of repeatedly computing the same graph paths, the system uses stored copies, reducing computational load and enabling horizontal scaling of machine learning models
2Productivity
If conventional graph database implementations are used with machine learning models, then real-time data consumption can be achieved, but horizontal scaling to support both real-time and batch processing is not possible
Solution Approach 1:
The patent separates graph processing into distinct segments: precomputation of graph paths (can be done in batch), caching of results, and real-time query execution. This segmentation allows different processing modes to coexist and scale independently, enabling horizontal scaling for both real-time and batch machine learning workloads
Solution Approach 2:
By precomputing and caching graph paths beforehand, the system enables flexible scaling during real-time operations. The cached data serves as a foundation that can support multiple concurrent machine learning models and processing tasks, enhancing adaptability and horizontal scaling capability
3Measurement precision
If graph paths are computed in real-time for dense vertices, then accurate graph attributes can be obtained, but computational resources become insufficient to meet SLA requirements
Solution Approach 1:
The patent precomputes graph paths and stores them in a cache before they are needed for machine learning processing. This eliminates real-time computation of graph paths, ensuring both accurate graph attributes are available and computation time requirements are met for SLA compliance
Solution Approach 2:
The system creates cached copies of precomputed graph path data that can be rapidly accessed during real-time processing. These copies provide accurate graph attributes without requiring expensive real-time calculations, meeting both precision and time requirements
Data Source
AI summary
A method for utilizing a graph path cache to facilitate real-time data consumption by a plurality of machine learning models is disclosed. The method includes receiving an input from a source, the input relating to a request to characterize a data element; retrieving a data attribute that corresponds to the data element from a data management system; determining, in real-time using the graph path cache, a graph attribute that corresponds to the data element by performing deep link analysis on a graph database; executing, in real-time, a model by using the data attribute and the graph attribute, the model corresponding to the request in the input; and transmitting, in real-time, a result of the executed model to the source in response to the input.