Word Extraction Device Combining Syntactic and Semantic Parsing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Lexical Knowledge Extraction (LKE) techniques can only generate extraction rules based on either syntactic or semantic information, limiting the accuracy of word extraction as they fail to consider both types of information simultaneously.
Innovation Solution
A word extraction device that combines syntactic and semantic parsing techniques to generate combined lexical representations, allowing for the extraction of both syntactic and semantic information from text, thereby improving the accuracy of word extraction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only syntactic parsing technique is used, then syntactic information can be extracted, but semantic information cannot be extracted
Solution Approach 1:
The patent combines syntactic parsing and semantic parsing techniques into a unified extraction rule generation system. The rule generation unit integrates rules from both parsing methods, allowing the system to simultaneously extract syntactic information (from dependency parsing) and semantic information (from semantic parsing), thereby resolving the contradiction between extraction accuracy and information coverage
2Measurement precision
If only semantic parsing technique is used, then semantic information can be extracted, but syntactic information cannot be extracted
Solution Approach 1:
The system merges semantic parsing rules with syntactic parsing rules in the rule generation unit. This combination enables the extraction system to access both semantic relationships (from semantic parsing) and syntactic structures (from dependency parsing), simultaneously improving extraction accuracy and expanding information coverage
3Reliability
If extraction rules are generated from training data, then the system can learn patterns, but it cannot generalize to new documents without additional processing
Solution Approach 1:
The system performs preliminary action by generating extraction rules from training data in advance. The rule generation unit creates a comprehensive rule set that captures patterns from multiple parsing techniques, which can then be applied to new documents without requiring retraining, thereby ensuring both reliability and adaptability
Data Source
AI summary
Aspects relate to improving accuracy of word extraction. A word extraction device includes a lexical representation generation unit for acquiring training data including sentences in which target extraction words are specified, generating a first lexical representation by processing the training data with a first parsing technique, generating a second lexical representation by processing the training data with a second parsing technique, and generating a first combined lexical representation by combining the first lexical representation and the second lexical representation; a query representation generation unit for generating, based on the first combined lexical representation, an extraction query representation indicating a query for extracting the target extraction words from a predetermined target search document; and a word extraction unit for extracting, by using the extraction query representation, extraction information that indicates information about the target extraction words from a second combined lexical representation generated based on the target search document.