Ontology-based Knowledge Acquirement, Management And Application

Today is the knowledge explosion era, knowledge and technology-intensive industrieswill replace the labor-intensive industries. In industrialization society, society's mainfunction has changed from goods production (fabrication) to knowledge economy. Sincethe computer scientist Professor Feigenbaum (BAFeigenbaum) from Stanford Universityhad put forward the concept of knowledge engineering in the fifth International Conferenceon Artificial Intelligence in1977, knowledge engineering has made a long-termdevelopment in management science and information science as a response for knowledgesociety and knowledge economy. Knowledge engineering takes knowledge engineers as themain body, it aims at solving application problems which depend on expert knowledge orexpertise, it carries out knowledge representation, knowledge inference and knowledgemanagement after knowledge acquisition oriented to experts or reliable resources ofknowledge, finally it solves real dependent issues by using knowledge effectively.Knowledge management is management science which takes the knowledge resources as"management" assets, it carries out scientific systematic management on organizationinformation and knowledge resources, it organizes knowledge effectively and creates newknowledge resources to enhance the efficiency of users' production activities. It shows thatthe relationship is closed between knowledge engineering and knowledge management;they are integrated in technology, including the collection and classification of knowledge,knowledge memory and tag technology, knowledge search technology, knowledgedissemination and use technology.The existing technologies of knowledge engineering and knowledge management aremostly based on grammar level which led reasoning difficult and knowledge applicationrestrained. It is because that knowledge representation lacks of the support of semanticsand the formalization is low, it can not reason automatically and makes the efficiency ofuse decline in exponential. In recent years, the emergence of semantic Web technologypromotes the development of knowledge representation based on semantic technology,Semantic Web is mainly based on XML, RDF/RDFS, OWL and so on, it constructsontology and logic reasoning rules based on semantic Web to carry out knowledge representation and knowledge reasoning based on semantic Web which makes computerscan understand and deal with Web information. The emergence of methods and tools ofontology description make the domain knowledge formalization and semantics sharingbecome possible.Aiming at practical application requirements, the current situation and existingproblems for the research field of knowledge engineering and knowledge management, thispaper put forward an application architecture named OKAMA based on ontology whichfocuses on knowledge acquisition, knowledge management and knowledge application.This architecture takes ontology as the core, and it takes semi-automatic knowledgeacquisition for Web-oriented text and reference books, knowledge management based onontology and knowledge inquiry and knowledge mining ontology based on semantic as themain technologies. This paper focus on the algorithm of semi-automatic knowledgeacquisition for Web-oriented, the algorithm of structured automatic processing for text, theconcept system, property system and axiom system of domain ontology, the calculation ofcorrelation and approximation of concept semantic and the methods of knowledge mining.Finally, this paper gives a prototype system named OKAMA and constructs knowledgebases and expert systems of cardiovascular disease.The main task of thesis is as follows: An ontology-based knowledge acquisition,management, and application architecture OKAMA are raised, and detailed thecomposition of the various functional modules. The biggest feature of the architecture isthat all used in semantic ontology support. Knowledge acquisition process is explained,which aims at Tools books and Web pages text. According to the characteristics of Web text,the concept of upper and lower access to methods are given,which is oriented Web pagetext.The experimental results (degree of support and reliability) analysis shows that theeffectiveness of the method, using RDF/RDFS as a constructor sub-ontology formalrepresentation. Knowledge acquisition methods are given, which is text-oriented toolsbooks, and focus on the design of the text structured algorithms and inversion of theknowledge-based digestion method of consistency checking.Ontology concepts ofcardiovascular disease and properties of architecture are given, and detailed the structure ofthe ontology axioms way; and then presented the semantics related degree and similaritycomputation which is ontology of cardiovascular-oriented. Since the premise of incomplete information, gray relationship analysis can measure the similarity between samples, so thispaper analyzes the relationship between gray-induced similarities measure of fuzzyclustering, which is proposed based on the analysis of the relationship between gray fuzzyclustering method. After analysis of the relationship between gray and the theory of nuclearsimilarity, then this paper derived a new nuclear depending on the relationship betweengray theory, which is the gray relationship nuclear. At the same time, a new grayrelationship measure is induced by the nuclear, which used theory of nuclear machine, andthen built a two-way channel between the both. From the implementation point of view,this paper researched the method and implementation of semantic-based knowledge basequery, proposed an ontology statistical correlation combining with the semantic correlationrelated to the association rule mining method. In this method, take association rule miningas the goal, first of all to establish the field of ontology, and integration of a more generalontology system support association rule mining, and then comprehensive consideration ofthe statistical relevance of ontology and semantic relevance of quantitative calculationrules of relevance. Finally, the application of objective and subjective interestingnessinterestingness constraints uninteresting rules generation.In this paper, we gain a set of methods of knowledge acquisition, knowledgemanagement and knowledge application based on ontology for other areas. The set ofmethods can be standardized so as to promote in the more areas.