Construction Of System Engineering Documents Based Domain Knowledge Base

In recent years,under the influence of the search engine giant Google's the"knowledge graph",many well-known domestic companies have launched their related products,such as Baidu's "zhixin-schemas",Sogou's "knowledge cube" etc.In this context,the various industries have begun to consider building their own domain knowledge base,which can provide comprehensive and intelligent support for business decisions.In this paper,we construct the knowledge base system based on the system engineering documents in HKBZ domain,and integrate the structured,semi-structured and unstructured data in the field.The domain knowledge of the system mainly includes the relationship between documents and documents(including citation and similarity),the relationship between documents and entities,the relationship between entities and entities,and other related knowledge.In this paper,the construction method of the domain knowledge base and the current research status are analyzed,and the basic links of the three knowledge bases,named entity recognition,entity relation extraction and knowledge representation are described in detail.This paper mainly has the following three contributions:(1)In the process of entity relation extraction of knowledge base construction,according to the characteristics of system engineering documents,puts forward four kinds of entities based on WAF theory of feature vector construction method,and experiments are carried out to verify the analysis using a variety of classification model,experimental results show that the method of feature vector to construct a reliable basis;(2)In the process of knowledge representation of knowledge base construction,this paper analyses and compares several main methods of knowledge representation,in order to make up the defect of knowledge representation based on ontology in sharing,reusability and interoperability,this paper proposes a new method of JsonSchema based on object oriented knowledge;(3)In the system implementation process,this paper designs and implements a prototype system of domain knowledge base,which has the advantages of high automation,high knowledge reuse rate,high query speed and high hit ratio.In addition,this paper constructs the domain knowledge base from the perspective of full automation,so as to put forward a complete set of HKBZ domain knowledge base construction scheme,the scheme also has some reference value and guiding significance for the construction of knowledge base in other fields.