Knowledge Graph Triples With Confidence Feedback Updating
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Building knowledge graphs is a time-consuming and often manual process, and existing systems do not provide adequate indication of relationship veracity, leading to potentially outdated or irrelevant information being provided to users.
Innovation Solution
A knowledge graph data structure utilizing triple data units with subunits for confidence, confidentiality, and temporal properties, combined with automated generation and feedback-based training to enhance accuracy and relevance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual processes are used to build knowledge graphs, then relationship veracity can be assessed, but the construction process is time-consuming and inefficient
Solution Approach 1:
The system performs preliminary actions by automatically generating initial knowledge graph data structures from documents before user interaction. The automated generation creates draft triple data units with preliminary confidence values, which are then refined through user feedback, combining automation efficiency with human verification for veracity.
Solution Approach 2:
The system implements feedback mechanisms where user interactions with the knowledge graph (such as correcting or confirming relationships) are captured and used to update confidence values. This feedback loop continuously improves relationship veracity assessment while maintaining automated construction efficiency.
2Loss of information
If existing knowledge graph systems are used, then information can be retrieved, but there is inadequate indication of relationship veracity leading to outdated or irrelevant information
Solution Approach 1:
The system applies local quality by attaching specific confidence values to individual triple data units within the knowledge graph. Each relationship is assessed and marked with its own veracity indicator, allowing users to evaluate the reliability of specific information rather than treating the entire knowledge graph uniformly.
Solution Approach 2:
The system changes parameters by introducing confidence values as a new parameter for each relationship in the knowledge graph. This parameter dynamically reflects the veracity of relationships based on multiple sources and user feedback, enabling the system to track and indicate information relevance and reliability.
3Productivity
If automated generation is used to construct knowledge graphs, then construction efficiency is improved, but adequate veracity assessment may be compromised
Solution Approach 1:
The system merges automated generation with user feedback mechanisms. The automated process efficiently constructs initial knowledge graphs from documents, while user feedback is integrated to verify and refine relationship veracity, combining the speed of automation with the reliability of human assessment.
Solution Approach 2:
The system implements self-service by automatically updating confidence values based on user feedback without requiring manual intervention for each correction. The system autonomously processes feedback, recalculates confidence metrics, and updates the knowledge graph structure, maintaining efficiency while improving veracity.
Data Source
AI summary
Various examples are directed to systems and methods for utilizing relationship data in a computing system. The computing system may extract first relationship data from a document and determine a first confidence value describing the first relationship data. The computing system may write the first relationship data to a knowledge graph data structure. The computing system may serve a first user interface page to a user computing device associated with a first user and receive feedback data describing an accuracy of the first relationship data. The computing system may modify a first confidence subunit of a triple data unit associated with the relationship to describe an updated confidence value based on the feedback data and a trust score of the first user.