Text Processing System for Entailment Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing textual entailment recognition technologies face challenges in accurately determining the entailment relationship between sentences, particularly when the coverage rate is high, as they fail to adequately consider structural differences and dependency relationships between sentences, leading to incorrect identification of entailment and failure to identify influencing sentence structures.
Innovation Solution
A text processing system that extracts common substructures and feature amounts based on dependency relationships between sentences, using these to determine the entailment relationship, thereby addressing the limitations of existing methods by focusing on structural differences and dependency relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If existing textual entailment recognition technologies use coverage rate to determine entailment relationship, then the determination process is simple, but the accuracy deteriorates when coverage rate is high because structural differences are not considered
Solution Approach 1:
The patent segments the sentence structure into common substructures and non-common substructures. By dividing the entailment determination into these components, the system can focus on the structural differences in non-common substructures while maintaining simplicity through the systematic segmentation approach.
Solution Approach 2:
The patent introduces a new dimension of analysis by examining dependency relationships between substructures, not just the presence or absence of substructures. This dimensional shift from binary presence/absence to relational dependency analysis improves accuracy without proportionally increasing complexity.
2Ease of operation
If existing technologies focus only on common substructures between sentences, then the analysis is straightforward, but the ability to identify influencing structural elements deteriorates
Solution Approach 1:
The patent extracts non-common substructures from the sentence comparison process and analyzes their dependency relationships separately. This extraction allows the system to maintain straightforward analysis of common substructures while additionally capturing the influential structural elements in non-common substructures.
Solution Approach 2:
The patent applies different analysis depths to different parts of the sentence structure. Common substructures are analyzed for their presence and basic relationships, while non-common substructures receive more detailed local quality analysis focusing on their dependency relationships and influence on entailment.
3Measurement precision
If the system analyzes dependency relationships between substructures, then the entailment determination accuracy improves, but the computational complexity increases
Solution Approach 1:
The system segments dependency relationship analysis to focus only on non-common substructures and their relationships with common substructures. This selective segmentation reduces the overall computational complexity while maintaining the accuracy benefits of dependency analysis where it matters most.
Solution Approach 2:
The patent applies dependency relationship analysis partially - only to the extent necessary for determining entailment. By focusing on critical dependency relationships rather than all possible relationships, the system achieves improved accuracy without the full computational cost of exhaustive analysis.
Data Source
AI summary
A text processing system that is able to appropriately determine textual entailment between sentences with high coverage is provided. The text processing system is configured to execute: processing of extracting a common substructure that is a partial structure of a same type, the partial structure being common to a first sentence and a second sentence and, based on the a structure representing the first sentence and a structure representing the second sentence; processing of extracting at least one of a feature amount representing a dependency relationship between the at least one common substructure in the first and second sentences and a feature amount representing a dependency relationship between the common substructure in the first and second sentences and a substructure different from the common substructure; and processing of determining an entailment relationship between the first sentence and the second sentence by using the extracted feature amount.