问答系统是下一代搜索引擎的发展方向，它能更好的理解问题的语义和用户的具体意向，而且与传统返回链接不同是，问答系统是返回一个确定的答案。随着互联网上RDF（Resource Description Framework）数据不断增多，越来越多的知识库将知识存储成RDF的结构化数据，这其实就是知识图谱。中国的知识图谱也在蓬勃发展，像百度、搜狗都引入了知识图谱来优化其搜索引擎。基于知识库的问答系统也开始应用知识图谱，将问答系统的知识库以知识图谱的形式存储，有利于提高问题的理解。 本文的主要工作是对用户输入的中文问句进行自然语言进行处理，其中包括中文分词和词性标注，然后将问题分成四种类型，针对每一种问题设计不同的SPARQL查询模板，根据问题的模板将其转换成完成的SPARQL语句，利用Virtuoso SPARQL Query Editor在线编辑器对DBpedia知识库的RDF数据进行查询，抽取问题的结果。

Question and Answering system, which can better understand the semantics of the problem and the user’s specific intentions, is the direction of the next generation of search engines. And unlike the traditional search engines, the Q&A system returns a definite answer instead of the links. With the increasing number of RDF （Resource Description Framework）data on the Internet, more and more knowledge base(KB) store knowledge into structured data(RDF), which is actually the knowledge graph(KG). Chinese KG is also flourishing, like Baidu and Sogou. The companies also have introduced KG to optimize their search engine. The Q&A system based on KB also began to use KG, which storage data in KG. It is a good way to improve the understanding of question. The main work of this paper is to deal with the Chinese question input by the user by natural language processing, including Chinese word segmentation and Part-of-Speech tagging. And then the work is to divide the question into four types. I need design different SPARQL query templates for each type of question. After that convert the question to the completed SPARQL statement. Then use the Virtuoso SPARQL Query Editor online to send query request to the DBpedia knowledge base( RDF). Finally extract the results of the question.