| This paper is supported by National High Technology Research and Development Program of China(Integrated Environment of Intelligent Applications, IEIA), our research area is knowledge sharing and knowledge exchange. We study on the algorithms of the exchange method between different knowledge representations and different uncertainty models. Our task is to implement the knowledge base visiting service of the knowledge base middleware of IEIA. It includes Rule/KIF and Bayesian/KIF knowledge interchanging.In recent years, with the development of computer applications and network technology, more and more knowledge resources appear in programs.Because the representations are not universal, we are not able to use the resources on the web and terminals effectively. How to organize and manage the knowledge and do the necessary exchanges to make the knowledge and information sharable and reusable becomes a hot research topic.Ontology was introduced to the domain of computer science, since knowledge engineering developing steadily. The advantages of using ontology may include: firstly, the analyses of ontologies clarify the structure of domain knowledge, thus make the foundation of knowledge representation; secondly, ontology is a kind of unified terminology and concept, thus make knowledge sharing possible. We always use ontology when we do knowledge exchange. Ontology is a concept model which can describe the system on a semantic and knowledge level. It can acquire knowledge from a specific field in a universal way, and provide the common understandings about the concepts in the field. We use standard ontology to exchange knowledge for not loosing the information of languages. We should design standard ontology. We need some objective standards when we think about how to design ontology to construct the concepts for knowledge exchange and reusing. We introduce five ontology design standards in our paper: Clarity, Coherence, Extensibility, Minimal Encoding Bias and Minimal Ontological Commitment. The standards can guide and evaluate our design and make sure that the ontology model can play a better role. An example is given to show the meanings and the relations of the five standards.Without the support of machine languages, computer can not transact knowledge. First Order Logic(FOL) is an important way to represent knowledge. Knowledge Interchange Format(KIF) is a language based on FOL. KIF can be used for the translation between different computer programs, it is built by Stanford University and it is a proposed draft American National Standard. KIF has been widely used in expert system, database and intelligent agents as a bridge between two knowledge representation languages. In order to make the knowledge sharing and knowledge exchange available and ease the development, many standards and protocols are brought out, the importance of KIF is more and more obvious.We have studied the translation method between KIF and Rule and Procedure Based Language(RPBL) and method between KIF and Bayesian Network Interchange Format(BNIF) in this paper. With the constructing of KIF ontology, we build a bridge between two different knowledge structures, providing ways of knowledge exchanging in expression level. XML serialization can make it easier to transact KIF ontology, and make the knowledge represented by KIF easy to use, transit and save, efficiently using in Web Service, so we define a basic KIF/XML representation method.This paper has also discussed the transformation between the Certainty Factor Model and Bayesian Model. These two models are the most used uncertain reasoning models in IEIA. During the applications of expert system, we usually do not have enough evidence to make decision, so there are many uncertain facts. How to handle these uncertainties needs uncertain reasoning model. In the last several decades, many researchers have tried many methods to represent and transact uncertainty, such as evidence theory, certainty factor model, PROSPECTER model and fuzzy set theory.In recent years, more and more people focus on probability method. If different expert systems use different uncertain reasoning models in a distributed expert system, it is necessary to transform the uncertainty of a proposition from one model to another when they cooperate to solve problems. In order to implement the transformation of uncertainties between the certainty factor model and the Bayesian model, we find that the key problem is how to obtainthe values of prior probabilities. This paper provides three methods to implement the transformation. They are expert value method, evidence instance method, synthesize instance method.Finally, as the application of the theory study result, the paper designs and implements the knowledge base transformation system between Bayesian knowledge base and rule knowledge base. These two knowledge base are the most frequently used in the expert systems of IEIA. The transformation architecture is composed of Bayesian knowledge base edit interface, knowledge exchange process and uncertainty factor model based rule knowledge base edit interface. Knowledge engineers can make use of human-machine interface to edit source knowledge base and target knowledge base; the transformation involves knowledge base edit, translation in knowledge structure, the transformation of uncertainty and knowledge base edit after interchanging. During the mapping process we need to finish two main tasks: the knowledge representation disposal and uncertainty disposal. We take KIF ontology as our exchange standard, follow the methods we told before, we change the inner knowledge structure from one form to another. In order to show the macroscopic view of knowledge sharing, we provide a knowledge sharing architecture which describes the interaction of five roles: ontology server, network clients, remote applications, knowledge based programs and agents. |
