Knowledge Graph Sequence Modeling With Subgraph-Based Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing knowledge graph management systems face inefficiencies in handling large amounts of data and lack accuracy in predictive models, leading to high resource consumption and unsatisfactory prediction outcomes.
Innovation Solution
The method involves extracting subgraphs from historical knowledge graphs, determining node and statistical features, and using a Temporal Graph Neural Network (TGNN) to train a prediction model based on these features, focusing on changes within the knowledge graphs to reduce data processing and enhance prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional methods process complete historical knowledge graphs for prediction, then comprehensive data analysis is achieved, but resource consumption increases and processing efficiency decreases
Solution Approach 1:
The patent segments the complete historical knowledge graph into multiple subgraphs based on temporal intervals. Each subgraph contains a portion of the historical data, allowing parallel processing and reducing the computational burden on a single processing unit. This segmentation enables efficient handling of large-scale knowledge graphs while maintaining comprehensive analysis capabilities.
Solution Approach 2:
The patent extracts only the necessary subgraphs from the complete historical knowledge graph that are relevant for prediction tasks. By taking out and processing only the essential portions rather than the entire graph, the system reduces resource consumption while maintaining prediction accuracy.
2Measurement precision
If traditional models analyze all historical knowledge graph data, then prediction accuracy is improved, but training time and computational resources increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing and segmenting the historical knowledge graph into subgraphs before the actual prediction task. This preliminary segmentation and organization of data allows the model to train more efficiently on structured, manageable portions rather than raw complete graphs, reducing training time while preserving accuracy.
Solution Approach 2:
The patent applies partial action by processing only the necessary subgraphs rather than the complete historical knowledge graph. This selective processing approach maintains prediction accuracy by focusing on relevant data portions while significantly reducing computational time and resources.
3Reliability
If the prediction model processes complete knowledge graphs, then comprehensive correlation analysis is achieved, but convergence speed decreases
Solution Approach 1:
The patent segments the knowledge graph processing into multiple subgraph analyses that can be performed in parallel or sequentially with reduced computational overhead. This segmentation maintains comprehensive correlation analysis by ensuring all relevant subgraphs are processed, while accelerating convergence through efficient resource utilization and parallel processing capabilities.
Data Source
AI summary
The present disclosure relates to a method, a device, and a program product for managing knowledge graphs. In a method, a historical sequence of knowledge graphs is obtained, wherein the historical sequence includes a plurality of historical knowledge graphs respectively associated with a plurality of historical time points, and a historical knowledge graph in the plurality of historical knowledge graphs includes a plurality of nodes and at least one edge between the plurality of nodes. Based on nodes and edges included in the plurality of historical knowledge graphs, a plurality of subgraphs respectively associated with the plurality of historical knowledge graphs are determined. A sequence feature associated with the historical sequence of the knowledge graphs is determined based on the plurality of subgraphs. A prediction model is trained based on the sequence feature such that the prediction model can describe the correlation among the plurality of historical knowledge graphs.