Research And Realization Of Question Analysis And Information Retrieval In Chinese Question Answering System

As Internet popularizes, more and more information are available online, providinga huge amount of resources for people. However, with the over whelming information,people find it very difficult to find the information they need quickly and accurately.There are some search engines that can help people to search for needed information,but they are far from being satisfactory. Question answering (QA) technology isdeveloped to meet people's demand. The inputs of search engines are combinations ofkeywords, while QA systems enable users to input a question in natural languages. Theoutputs of search engines are some related web pages, while QA systems can give usersconcise and accurate answers. By using QA systems, users can find the information theyneed conveniently, quickly and accurately.QA technology is a very hot research direction in the field of natural languageprocessing (NLP). It combines a large variety of NLP technologies. Presently, manyresearch institutes have been investigating English QA technologies. Some matureEnglish QA systems have been widely recognized. But few institutes are doing researchon Chinese QA systems. No Chinese QA systems have been proposed up to now. In thispaper, we try to investigate some technologies for Chinese QA systems.Our Chinese QA system called Virtual Information Consultant. The VIC system hasfive main parts: Question Analysis, Information Retrieval, Information Processing,Answer Extraction, and Frequently Asked Questions Module. Question Analysis is theprimary task of VIC System. The result of Question Analysis has a great effect onfollowing processing work, even on finding the correct answer. It 's also the mostimportant research aspect on Intelligent consulting System, Man-machine dialogue, andso on.In this paper, we research deeply on Question Analysis, considering the character ofChinese and technique of computational linguistics. We classify query words detailed,and present standard query sentence that enable VIC system use the same algorithm toprocess the query in different format. Via identifying topic word, the destination sourcecan be found that can decrease searching of no use information. The paper also gives thedesign of the next phase, Information Retrieval, combining with the trait of QuestionAnalysis.Finally, we do the tests and evaluations of the technique applied to QuestionAnalysis. The results indicate the rationality and reliability of the theories in this paper.