Research Of Ontology-based Component Description And Intelligent Retrieval

With the expansion of computer applications, whether in manufacturing, e-commerce or the field of the service consulting, computer software has become an important infrastructure and the foundation to promote the development of the new economy in modern society. However, the traditional software development methods have been unable to meet the requirements of society, software reuse is a effective method that is proposed for improving the quality of the software and the efficiency of software development. Component-based software development is an effective form of support software reuse, and gradually become an effective way to solve the software crisis, improve software productivity and ensure software quality. However, with the growing size of the component library, it is becoming more and more difficult to retrieve the components which can meet the demand of the reuse of the components in the component library.Effective description of the component is the permise of the storage and retrieval of the component. Classification and description based on facet is one of the more widely in the actual application method, but the efficiency of the retrieval is not high because of the matching is based on syntax level which is lack of capabilities in representation, understanding and pro-cessing of the knowledge. To solve this problem, the key is to make the component retrieval from syntactic matching based on simple search keywords up to the semantic matching based on knowledge. Ontology based on facet classification description is an effective solution which can increase the sharing and reuse of the components, and can leverage its reasoning logic to realize the efficiency of component retrieval. Howerver, the ontology-based component description improves the semantic of component retrieval, but with the expansion of the scale of the component library, the increasing complexity and semantic of component, it is difficult for users to determine the search conditions, and it is difficult to find the desired component quickly from a large number of search results. To deal with these problems, we proposed a retrieval solution based on Multi-Agent. We established a Multi-Agent System (MAS) which is autonomous and cognitive by creating various of Agent and achievied a component retrieval system with flexibility, accuracy and intelligence. Fianlly, in the study of ontology-based component description and MAS, we implemented the prototype system of component retrieval based on ontology and MAS. The feasibility of the theory through the prototype system proves that component retrieval base on ontology and Multi-Agent has higher recall, precision and query efficiency.