Intelligent Question Answering System Using Semantic Analysis and Modular Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing intelligent question answering systems have low coverage and provide non-deterministic, inaccurate answers, especially for moderately to less frequently and rarely asked questions.
Innovation Solution
A method and system that perform semantic analysis of questions and distribute them to corresponding search processing systems, including semantic relationship mining, text library search, knowledge base search, and question and answer library search systems, to broaden the search scope and improve answer accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing intelligent QA systems rely on QA site community data, then the system can return similar answers, but the coverage of questions is low and accuracy is insufficient for moderately to rarely asked questions
Solution Approach 1:
The patent divides the QA system into multiple independent search modules: semantic relationship mining system, text library search system, knowledge base search system, and QA library search system. Each module handles different types of questions independently, allowing the system to maintain high accuracy for common questions while expanding coverage to rare questions through specialized modules.
Solution Approach 2:
The patent creates a universal QA system that can handle multiple types of questions through different search modules. The semantic relationship mining system handles relational questions, the text library search system handles factual questions, the knowledge base search system handles domain-specific questions, and the QA library search system handles community questions, making the system adaptable to diverse question types.
2Measurement precision
If the system uses multiple search processing systems, then the search scope is broadened and answer accuracy is improved, but the system complexity increases
Solution Approach 1:
The patent segments the search function into four distinct modules, each responsible for specific types of questions. This segmentation allows the system to achieve high accuracy by specialized processing while managing complexity through modular architecture, where each module can be independently optimized and maintained.
Solution Approach 2:
The patent introduces a question analysis system as an intermediary that receives user questions, performs semantic analysis, and routes them to the appropriate search module. This intermediary layer simplifies the overall system architecture by providing a single entry point and intelligent routing mechanism, reducing the complexity of handling multiple search systems directly.
Data Source
AI summary
Embodiments of the present invention provide a method, system and storage medium for implementing intelligent question answering. The method includes: receiving a query question; performing a semantic analysis of the question; performing corresponding search processing for the question based on a result of the semantic analysis, wherein the search processing includes search processing performed for the question by at least one of a semantic relationship mining system, a text library search system, a knowledge base search system, and a question and answer library search system; and returning an answer based on a result of the search processing. In this way, the accuracy of answers to the questions is effectively improved.