Research On Automatic Question Answering System Of Tourist Area Based On Semantic Web

With the increasing amount of information on Internet, the demand of user’s message retrieval become more and more specific. People are more likely to present their question in natural language. Automatic question answering system was born in that background. It will parse natural language questions and then check out specific answers from Internet and knowledge base.This kind of demand is quite obvious in tourist. This article aims to design an efficient automatic answering system applied in tourist to make natural language understood by computer, which may improve user’s experience in message retrieval.In this article, the conception of ontology and semantic web is introduced at first. The whole structure of the automatic answering system is represented after that. Then, this article presents the improvement of word segmentation algorithm. After that, it presents the plan in which keywords in natural language be taken out by regular expression. This article also designs methods to generate SPARQL by Stanford Parse, which is a syntax parsing tool produced by Stanford University. At last, it designs the procedure engine based on state machine in order to keep the context dependence, so that users could make their requests more frequently.