Self-Evolving Knowledge Graph Automation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional knowledge graph construction and updating require manual effort from subject matter experts, making the process complex and time-consuming, especially when new documents are published, necessitating continuous updates to reflect changing relationships between entities.
Innovation Solution
A self-evolving knowledge graph system that automatically updates by processing new documents using a vector space model, calculating relevance values, and updating confidence scores based on cosine similarity thresholds, allowing for scheduled updates that match publication schedules, thereby reducing manual intervention and ensuring the knowledge graph remains up-to-date.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual assembly and updating of knowledge graphs is performed by subject matter experts, then accuracy and reliability of entity associations are improved, but time consumption and operational complexity increase significantly
Solution Approach 1:
The system enables self-service by automatically processing new documents, extracting entity associations, and updating the knowledge graph without requiring manual intervention from subject matter experts. The automated pipeline includes document ingestion, entity recognition, association extraction, and confidence score calculation, allowing the knowledge graph to evolve autonomously while maintaining reliability through multiple validation mechanisms
Solution Approach 2:
The patent replaces the mechanical manual process of expert review and curation with an automated computational system. Machine learning models and natural language processing algorithms substitute for human experts in analyzing documents and identifying entity associations, dramatically reducing time consumption while maintaining association quality through algorithmic validation
2Reliability
If continuous manual updates are performed to reflect new publications, then knowledge graph currency and relevance are improved, but operational complexity and resource requirements increase
Solution Approach 1:
The system implements periodic action by scheduling automated updates at predetermined intervals or triggered by new document availability. The update process operates in discrete batches, ingesting new publications and refreshing the knowledge graph periodically rather than requiring continuous manual monitoring, thereby maintaining currency while reducing operational complexity
Solution Approach 2:
The knowledge graph system performs self-updates by automatically detecting new documents, processing them through the extraction pipeline, and integrating findings without external intervention. This autonomous operation eliminates the need for complex manual coordination while ensuring the knowledge graph remains current with newly published information
3Loss of information
If comprehensive document processing is performed to ensure complete coverage of entity associations, then information completeness is improved, but computational requirements and processing time increase
Solution Approach 1:
The system applies local quality by focusing computational resources on processing only the portions of documents that contain entity associations and relevant relationships. Rather than uniformly processing entire documents, the system identifies and extracts only the necessary information segments, maintaining completeness of entity associations while reducing overall computational requirements through selective processing
Solution Approach 2:
The patent employs partial action by processing a subset of documents or document portions that are most likely to contain valuable entity associations. The system uses filtering mechanisms and confidence score thresholds to identify high-value content, achieving sufficient information completeness without the excessive computational cost of processing every detail of every document
4Productivity
If automated processing is implemented to reduce manual effort, then productivity is improved, but measurement precision and reliability of entity associations may deteriorate
Solution Approach 1:
The system implements feedback mechanisms by calculating confidence scores for each extracted entity association and using these scores to validate and filter results. The automated pipeline includes validation steps that compare extracted associations against existing knowledge graph data and apply quality thresholds, ensuring that productivity gains from automation do not compromise the precision and reliability of detected relationships
Data Source
AI summary
A computer system updates a knowledge graph. A model corresponding to a set of documents is received, wherein the model comprises a plurality of entities, a plurality of entity associations, and a plurality of confidence scores corresponding to the plurality of entity associations. A relevance value is calculated for each entity of the plurality of entities that are present in the set of documents and for each entity of the plurality of entities that are present in a new document. One or more entity associations that are supported by specific portions of the new document are identified. The confidence scores for each of the identified one or more entity associations are updated based on a level of support in the new document. Embodiments of the present invention further include a method and program product for updating a knowledge graph in substantially the same manner described above.