Knowledge Graph Entity Incorporation via Candidate Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for incorporating entities into knowledge graphs are time-consuming and less accurate due to the need for entity resolution on all existing entities, which is not feasible for large datasets like tens of billions of entities, and use methods like cosine similarity that are not precise.
Innovation Solution
A method that selects a candidate entity associated with the to-be-incorporated structured entity from a knowledge graph, determines the associated entity based on prior attribute information and a preset model, and merges this entity into the graph, reducing the number of compared entities and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If entity resolution is performed on all entities in the knowledge graph, then entity incorporation completeness is improved, but computational time and resource consumption increase significantly
Solution Approach 1:
The patent segments the entity resolution process into two distinct stages: (1) candidate entity selection that filters entities based on category prior attribute information, and (2) precise matching using preset models only on selected candidates. This segmentation reduces the scope of computationally intensive operations while maintaining incorporation completeness.
Solution Approach 2:
The patent performs preliminary filtering by selecting candidate entities based on category prior attribute information before conducting the actual matching process. This preliminary action eliminates irrelevant entities early, reducing the computational burden of subsequent precise matching operations.
2Ease of manufacture
If cosine similarity or BOW vector methods are used for entity matching, then computational simplicity is improved, but matching accuracy deteriorates
Solution Approach 1:
The patent applies different matching strategies to different parts of the entity matching process: category prior attribute information is used for candidate selection, while preset models (which may include more sophisticated algorithms beyond simple cosine similarity) are applied to the selected candidates. This local quality approach optimizes both simplicity and accuracy in different stages.
Solution Approach 2:
By performing preliminary filtering based on category attributes before applying the final matching model, the system prepares the data in a way that enhances the effectiveness and accuracy of the subsequent matching operation, rather than relying solely on simple cosine similarity across all entities.
Data Source
AI summary
Embodiments of the disclosure disclose a method, apparatus, server, and storage medium for incorporating a structured entity, wherein the method for incorporating a structured entity can comprise: selecting a candidate entity associated with a to-be-incorporated structured entity from a knowledge graph, determining the to-be-incorporated structured entity being an associated entity based on prior attribute information of a category of the candidate entity and a preset model, merging the associated entity and the candidate entity, and incorporating the associated entity into the knowledge graph. The embodiments can select a candidate entity, and then integrate a preset model using prior knowledge, which can effectively improve the efficiency and accuracy in associating entities, and reduce the amount of calculation, to enable the structured entity to be simply and efficiently incorporated into the knowledge graph.