Font Size: a A A

Chinese Auto Question-Answering System For Computer Domain

Posted on:2007-03-03Degree:MasterType:Thesis
Country:ChinaCandidate:H L ZhangFull Text:PDF
GTID:2178360212980090Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
As Internet popularizes, more and more information are available online. The development of search engines helps people to search information conveniently: Users need input keywords, and then search engines return some related web pages. But with the overwhelming information, users find it very difficult to find the information they need quickly. Therefore, they put forward the higher request: users hope they can find the information quickly, accurately and exhaustively. Question answering (QA) system is developed to meet people's demand.QA system combines a large variety of natural language processing (NLP) technologies. It involves several courses, such as calculation linguistics, information science and artificial intelligence. So QA system is among the very hot research direction in the field of computer application. In QA system, users can ask questions in natural languages, system will give users concise and accurate answers. So users can find the information they need conveniently, quickly and accurately.According to the character of computer knowledge, this paper design and realize a Computer_domain automatic Chinese QA system based frequently-asked question (FAQ). Our Chinese QA system has three main parts: word division,question analysis and FAQ. The word division is the pivotal technology of NLP. The quality of word division affects directly to the comprehension degree of user's question; Question analysis is used to clarify the intention of user's question. Question analysis includes question classification, keyword selection and keyword extension; FAQ stores the most frequently asked questions and their answers. For user's questions, QA system can search question set in FAQ, find the similar question according to semantic similarity for sentence, and give the most relevant answer to user. In our system, the size of FAQ is very important, which affect performance of the whole system. So, we design practical interface to update and maintain FAQ.At last this paper introduces the realization of out QA system and the evaluation of the system.
Keywords/Search Tags:Frequently-asked Question (FAQ), Natural Language Processing (NLP), Semantic Similarity, Question Answering
PDF Full Text Request
Related items