Dialogue System Using Question-Answering for Semantic Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current dialogue systems struggle to engage in natural interactions with users due to limitations in understanding semantic relations and dependencies in user utterances, leading to irrelevant responses and inability to control the intention of system utterances across various topics.
Innovation Solution
A dialogue system that includes a parameter generating mechanism to analyze user input, a question-answering system to obtain answers, and a response generating mechanism to produce relevant responses based on these answers, utilizing morphological and syntactic analysis, factoid and non-factoid question generation, and utterance shaping to ensure natural dialogue flow.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If keyword search is used to generate system utterances, then the system can retrieve information from web documents, but the search results contain excessive unnecessary information making it difficult to extract essential information
Solution Approach 1:
The patent extracts only the essential information from web documents by using a question-answering system that specifically targets and retrieves answers to generated questions, rather than presenting all search results. This filters out unnecessary information while preserving key content.
Solution Approach 2:
The system changes the parameter of information retrieval from broad keyword-based search to targeted question-answering. By formulating specific questions based on user utterances and extracting precise answers, the system transforms the retrieval process to focus on essential information only.
2Adaptability or versatility
If scripts with predetermined dialogue rules are used, then the system can generate responses based on extracted words, but the number of response patterns is limited making natural interaction impossible
Solution Approach 1:
The system uses a question-answering mechanism that automatically generates appropriate responses based on user utterances without requiring pre-prepared scripts for each possible dialogue scenario. The system serves itself by dynamically creating and answering questions relevant to the conversation context.
Solution Approach 2:
The dialogue system transitions from static predetermined scripts to dynamic question-generation and answer-extraction. The system adapts its responses based on the actual user input by generating context-specific questions and retrieving relevant answers, making the dialogue flexible and natural.
3Reliability
If semantic relations of user utterances are not correctly understood, then the system can process user input, but irrelevant system utterances are generated breaking natural dialogue flow
Solution Approach 1:
The patent introduces a question-answering system as an intermediary between user utterance processing and response generation. This intermediary generates appropriate questions based on user input and retrieves accurate answers, ensuring that the semantic relations are correctly understood and reflected in the system responses.
Data Source
AI summary
A dialogue system includes: a question generating unit receiving an input sentence from a user and generating a question using an expression included in the input sentence, by using a dependency relation; an answer obtaining unit inputting the question generated by the question generating unit to a question-answering system and obtaining an answer to the question from question-answering system; and an utterance generating unit for generating an output sentence to the input sentence, based on the answer obtained by the answer obtaining unit.