Knowledge Graph Instance Fusion via Iterative Relation Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for fusing different instances describing the same entity in knowledge maps are inadequate, leading to redundancy and errors due to reliance on attribute similarity, which is difficult to set perfectly, resulting in incorrect identification of instance pairs.
Innovation Solution
A method and device that utilize a connection diagram to identify and fuse instances based on instance relations, iteratively updating the diagram by adding connection lines until a specified condition is met, expanding instance relations, and calculating correlations and similarities to accurately identify equivalent instances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If attribute similarity is used as the only criterion for instance fusion, then the fusion process becomes simple, but the identification accuracy of instance pairs describing the same entity deteriorates
Solution Approach 1:
The patent changes the parameters used for instance fusion from only attribute similarity to a comprehensive set including instance relations, attribute similarities, and correlation coefficients. This resolves the contradiction by expanding the parameter space to achieve both operational feasibility and identification accuracy.
Solution Approach 2:
The patent creates a composite evaluation framework that combines multiple factors (instance relations, attribute similarities, correlation coefficients) to assess instance equivalence. This composite approach overcomes the limitation of using a single criterion, achieving both simplicity and accuracy.
2Measurement precision
If attribute fuzzy matching rules are set perfectly to improve instance identification accuracy, then the fusion precision improves, but the device complexity and difficulty of implementation increase
Solution Approach 1:
The patent segments the instance fusion process into distinct modules: instance relation extraction, attribute similarity calculation, correlation coefficient computation, and comprehensive evaluation. This segmentation reduces the complexity of implementing perfect matching rules by breaking them into manageable, independent components.
Solution Approach 2:
The patent implements a feedback mechanism where the system iteratively refines instance pair identification based on correlation coefficients and instance relations. This feedback loop allows the system to automatically adjust and improve identification accuracy without requiring manually configured complex fuzzy matching rules.
3Measurement precision
If iterative fusion and expansion operations are performed to adequately identify equivalent instances, then the instance pair identification accuracy improves, but the processing time and productivity deteriorate
Solution Approach 1:
The patent performs preliminary actions by pre-computing and storing instance relations and attribute similarities before the fusion process. This preparation reduces the computational burden during iterative operations, maintaining high identification accuracy while improving processing speed.
Solution Approach 2:
The patent implements periodic action through iterative fusion and expansion operations that alternate between identifying instance pairs and updating the connection diagram. This periodic approach systematically improves identification accuracy while controlling processing time through structured iteration.
Data Source
AI summary
Disclosed is a method, a device and equipment for fusing different instances describing the same entity. The method includes: acquiring a connection diagram comprising a plurality of instances, where different nodes in the connection diagram represent different instances, and connection lines between the nodes represent instance relations between the instances corresponding to the nodes; based on the instance relations, identifying different instances describing the same entity in the connection diagram, fusing the nodes corresponding to the identified instances, and updating the connection diagram.