Entity Relationship Graph Construction via Grammatical Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for extracting knowledge from unstructured text data fail to accurately identify and represent entities and their relationships, resulting in ineffective image summarization and limited user benefit.
Innovation Solution
A method and apparatus for constructing an entity relationship graph, which involves entity recognition, grammatical structure analysis, vector conversion, and knowledge graph construction to improve the accuracy of text entity relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional information retrieval methods are used to extract knowledge from unstructured text, then the process is simple and fast, but the accuracy of entity relationship identification is poor and entities cannot be properly discovered
Solution Approach 1:
The patent segments the entity relationship extraction process into distinct modules: entity recognition module extracts entities from text, grammatical relationship graph construction module builds structural relationships, and entity relationship analysis module identifies semantic relationships. This segmentation allows each module to specialize in specific tasks, improving overall accuracy while managing complexity through modular design.
Solution Approach 2:
The patent introduces a grammatical relationship graph as an intermediary structure between raw text and final entity relationships. This graph serves as a mediator that captures grammatical dependencies and facilitates the transition from unstructured text to structured knowledge, improving entity relationship identification accuracy without directly handling the full complexity of unstructured data.
2Measurement precision
If entity recognition is performed on all text entities without unification, then the extraction process is simple, but duplicate entities affect the accuracy of the entity relationship graph
Solution Approach 1:
The patent performs entity unification as a preliminary action before constructing the entity relationship graph. By identifying and merging duplicate entities (e.g., resolving references like 'he' or 'she' to their full entity names) in advance, the system ensures that the subsequent graph construction operates on a clean, deduplicated entity set, thereby improving graph accuracy without significantly impacting overall extraction efficiency.
3Loss of information
If only flat document views are generated from text, then the summarization process is simple and fast, but the entities and their relationships cannot be discovered
Solution Approach 1:
The patent transitions from two-dimensional flat document views to a multi-dimensional knowledge graph structure. The knowledge graph adds dimensions for entity types, relationship types, and grammatical dependencies, enabling comprehensive representation of entity relationships that flat summaries cannot capture. This dimensional expansion preserves complete entity relationship information while organizing it in a structured, queryable format.
Data Source
AI summary
The present disclosure relates to the technical field of information retrieval, and discloses a method for constructing an entity relationship graph, including: performing entity recognition and an operation of uniting like terms on the text to be processed to obtain a standard text entity set; performing grammatical structure analysis on the text to be processed, to obtain a grammatical relationship graph; selecting the node connected to the connection edge in the grammatical relationship graph as a target connection node and performing vector conversion to obtain a text vector, and analyzing a text entity relationship between the target connection nodes according to the text vectors; constructing a knowledge graph according to the standard text entities and the text entity relationship to obtain the entity relationship graph corresponding to the text to be processed. The present disclosure can improve the accuracy of the text entity relationship graph.