A Representation And Management Method Of Ontology Changes

In 2000, Tim Berners-Lee firtst introduced the concept of Semantic Web. And it have a positive effect in solving the problem such as the information retrieval and the semantic procession we often confront in the world wide web. The main idea of the Semantic Web is to, by expanding the world wide web, have the information in the web carry semantic meanings. Therefore it enable the computers to better understand and process such information and finally facilitate the communications between users and terminals. This significant concept blazes a new trail of thinking for the problems in the current internet.As concerning the ontology, foundation of the next-generation semantic web, most of the researches focus on the problem of how to structure the ontology. The problem of how to manage the changes of the ontology has been rarely touched. However, the ontology, as a conceptive model with semantic inference ability, is constantly on the change. The changes of the ontology include the changes of its domain, the changes of its representative form, and the changes of its conceptive model. These changes might give rise to the consistency problem in between the applications, the service based on the ontology, and the ontology itself. Therefore how to maintain the consistency of the ontology while handling such changes has always been the main concern of the management of ontology changes. Most existing development tools are not formalized enough to deal with the such changes of the ontology. They still largely depend on the participation of humans. And, there still lacks a satisfactory criterion to value the change management strategies. As for this problem concerning the manengement of ontology changes, I think we may use the relatively mature methods and theories of other research fields for our reference.This article starts with a brief account of the basic concepts of ontology, the changes of ontology, and the application of the ontology in the computer industry. It then analyses the six phases of the changes of ontology, namely the capture phase, the representation phase, the semantic phase, the transmission phase, the implementation phase and the confirmation phase, of which it put an emphasis on the semantic phase. After the introduction of some relevant knowledge in the study of the belief revision, it makes a comparison between this study and that of the management of ontology changes. Then the article continues with the discussion about the feasibility of applying the theories and algorithms, now in general use in the study of the belief revision into the problem of how to manage the ontology changes. After analysing the features of ontology changes, it comes out with a simplified modification operation on the ontology. It then introduces the concept of the Set of Preliminary Ontology Changes and gives a proof for the completeness and minimality of this set. In this article it applies the principles of the belief revision and the relevant knowledge such as the epitemic entrenchment into the design of the management methods. When deal with a ontology change,we define a consistency constraints set that must be complied,and we also define the ontology changes management strategies that can be selected. The author also defines a new term-the entity reliability, and by the quantification of the entity reliability, the quantification of the whole process of the change of ontology could be realized. It works out with a new method of the management of ontology changes, which will select the optimal strategy that would restrict the of ontology changes to a minimum degree.Finally, it implemented this method, and by analysing experimental results, it reaches the conclusion that during the quantification of the process of ontology changes, it could get the optimal strategy which causes a minimum change in the ontology.And in the summary, it also points out the parts that will still be needing improvement.The author belives that since the study of the belief revision is relatively mature research field., many of its classic theories and algorithms could be adopted in the solving the problem of how to manage the changes of ontology. And this will be a promissing direction in the furture.