Automatic question-answering system based on theme-rheme positions and realization method of automatic question answering system

Abstract

The invention discloses an automatic question-answering system based on theme-rheme positions and a realization method of automatic question answering system. The automatic question-answering system comprises a preprocessing module, a question understanding module, an information retrieval module, an answer extracting module and a knowledge base module, wherein the preprocessing module is used for tagging participles and parts of speech in user questions; the question understanding module is used for classifying the questions and dividing the theme-rheme positions; the information retrieval module is used for performing answer retrieval and similarity sequencing; the answer extracting module is used for performing component extraction and answer synthesis on retrieved answers; the knowledge base module is used for constructing an answer language knowledge base. A similarity searching and matching method for answers of the question-answering system based on the theme-rheme positions on the basis of semantic comprehension is provided. Firstly, sections and chapters in the answer language knowledge base are divided into minor sentences; the minor sentences are divided into theme positions and rheme positions to express chapter structures of the minor sentences; the theme positions and rheme positions of the questions are analyzed to understand the question intentions of the user and learn known information of the user; unknown contents of the user are found from the answers so as to improve the accuracy of automatic question answering.

Description

technical field [0001] The invention relates to an automatic question answering system based on main rhetoric and its realization method. Background technique [0002] With the rapid development of Internet technology, massive amounts of information are stored on the Internet every day, and people continue to use search engines to obtain various types of information. However, the current search engines have great limitations. For example, users usually return a large number of related web pages, which are complicated and complicated. The automatic question answering system allows users to ask questions in natural language, understands users' questions through various natural language processing technologies, and returns more concise answers. [0003] However, as far as the current automatic question answering technology is concerned, most automatic question answering systems still search for the information needed by users based on keyword matching, and cannot understand th...

AI Technical Summary

Problems solved by technology

[0003] However, as far as the current automatic question answering technology is concerned, most automatic question answering systems are still based on keyword matching to search for the information users need, and cannot understand the real needs of users.

Sometimes, people's need to find information may not be able to express their intentions through simple keyword combinations. In this way, the automatic question answering system is even more unacceptable and understands the user's intentions, and then gives satisfactory answers.

Moreover, this form of keyword matching, using the simplest form of keyword matching to find relevant information, cannot deal with the deep semantics of users, so sometimes the retrieved answers vary greatly, and the accuracy rate is very low

Images