Question Answering System Using Similar Question Generation and Extension Word Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing question answering systems face challenges in providing highly accurate answers efficiently, especially when dealing with typical question patterns, as they require manual specification of Web pages containing answers and are prone to errors due to disuse, content changes, or URL changes.
Innovation Solution
A question answering system that includes a similar-question generation unit, a first search unit, a second search unit, and an answer selection unit, which selects a feature word from an input question, generates a transformational question, executes multiple search processes, and analyzes correspondence relations to determine the answer, thereby automatically selecting a Web page containing the answer without errors from changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual specification of Web pages containing answers is used, then answering accuracy for typical question patterns is improved, but maintenance cost and complexity increase due to disuse, content changes, or URL changes
Solution Approach 1:
The system automatically detects typical question patterns and selects appropriate Web pages without manual intervention. The automatic Web page selection unit monitors and updates Web page selections based on detected patterns, eliminating the need for manual maintenance and reducing complexity while maintaining accuracy.
Solution Approach 2:
The system pre-detects typical question patterns from multiple Web pages and stores them in advance. When a question is received, the system can quickly match it against pre-detected patterns and retrieve answers without requiring real-time manual analysis, improving both accuracy and efficiency.
2Reliability
If multiple search processes are executed to improve answer accuracy, then answering reliability is improved, but processing time increases
Solution Approach 1:
The search process is divided into two stages: first, detect typical question patterns from multiple Web pages to identify candidate answers; second, verify and select the most appropriate answer by comparing against the original question. This segmentation allows comprehensive searching while maintaining efficiency through structured processing.
Solution Approach 2:
The system performs searches on multiple Web pages (excessive action) but only processes and displays the most relevant results. By executing comprehensive searches initially and then filtering results, the system ensures high reliability while presenting only the necessary information to the user.
Data Source
AI summary
A question answering system includes a similar-question generation unit, a first search unit, a second search unit, and an answer selection unit. The similar-question generation unit selects a feature word from an input question to generate a transformational question in which the feature word is replaced with another word. The first search unit executes a first search process based on the transformational question. The second search unit selects a word included in a first search result acquired as an extension word and executes a second search process with using the selected extension word. The answer selection unit analyzes a correspondence relation among extension words included in a second search result, determines the feature word and a constructive part of an answer equivalent word in the second search result based on an analysis result, and selects a word determined to be an answer.