Design & Implementation Of A Large-scale & Real-time Knowledge Base (HKBE)

To fulfill the simulation of humankind intelligence,to fabricate a kind of machine which can substitute for human brain,both are the greatest dreams of human being. The emergence of AI and ES is just a good attempt to achieve the goal in this field. The significance of research on knowledge base lays on the theoretically directive effect on the design and research of intelligent computer.HKBE(Hohai Knowledge Base Development Environment) as a knowledge base development environment,can be called ES shell. With the need of practical ES,HKBE focuses on the following two aspects.Firstly,the paper introduces how HKBE employs some technologies of database to augment the body of knowledge base,and achieves a large-scale knowledge. In detail,HKBE takes use of RDB to represent,store and maintain production rules,converts the structureless production rules into a linear structure and stores the latter in RDB. In order to maintain the rules conveniently,we make use of the technology of RDBMS in managing data,consider the trait of production rules,and then design and realize the arithmetic to check the rules' equipollence and circulation on the basis of our former work-knowledge ordering. All these solve the knowledge base's consistency effectively.Secondly,the paper focuses on the efficiency of inference engine. Considering the fact that RDB stores the knowledge,we use the following steps:the relational construction and granularity disassembling of knowledge,implicational inference. Beyond the above,we consider the advantages of RDB and unite the technology of database's index,association and filter to project a mixed inference mechanism. The techniques mentioned advance the efficiency of inference engine highly,and proved availably.The paper also introduces the concept of knowledge representation,basic principles of knowledge base,some commonly used methods of knowledge representation and knowledge base system.