Research On The Method Of Knowledge Extraction And Knowledge Base Construction From Hudong Encyclopedia

In recent years, the rapid development of Internet brings great convenience to people’s life, as a result, the amount of information contained in itself was explosive growth. Faced with such a huge source of information, the way people access information has not fundamentally changed. The reason of this situation is that most knowledge is stored in webpages instead of knowledge. It is difficult for computer to understand. In order to make better use of the Internet, a lot of people build a large number of knowledge bases, with many different methods.Based on the background above, this paper proposes a method to build knowledgebase automatically. Using this method, this paper implements a complete system to build knowledge base from Hudong encyclopedia and visualize the search result of knowledge base with graph. Generally speaking, the main contribution of this paper described as follows.Firstly, in order to get the page of Hudong encyclopedia, this paper implements a crawler to obtain articles from Hudong encyclopedia walking through category webpages of Hudong encyclopedia.Secondly, this paper implement a system to extract structured information from these articles. In order to make the best of articles of Hudong encyclopedia, this paper proposes a Chinese-oriented open domain entity relation extraction method. After that, this paper does experiments to illustrate the effectiveness of this method.Thirdly, this paper implements a system to store knowledge extracted from Hudong encyclopedia as RDF format.The knowledgebase constructed in this paper contains a large number of entity relation triples. In order to provide a convenient user interface for knowledgebase, a visualization system is implemented. It allows users to retrieve the knowledgebase with one or two arguments in entity relation triples and show the result in the form of graphs.