The Research Of Component Library Management And Component Retrieval Method Based On Domain

As an effective approach to solve software crisis, software reuse is getting more and more attention, and with perfection of component specification such as CORBA, OpenDoc, OLE2 and JavaBeans, reuse emphases divert to how to develop reusable components, how to manage components and how to improve the component search efficiency. Software component library is important support in implement of software reuse. Whether or not succeed in software reuse to a great extent depends on the structure, elements and management of reuse component library.In the article, on the basis of domain knowledge, the domain models of specific domain are acquired, and components are developed, described, classified and managed based on the domain models, thus it forms the library model of managing components according to the domain. In management of component library, retrieval and selection of the components that meet the user's needs is the core questions. The search efficiencyaffects components' reuse directly, and the library organizes the components according to the domain, so the retrieval of component is limited in the specific domain library, which simplifies the search of the component to a certain extent. To help user fast locate the needed components, the article compares kinds of retrieval tactics of components, integrates based_on facet, key_word and full_text retrieval toghther, and proposes a new retrieval tactics. Compared with single component retrieval mechanism, it takes a great deal of advantages, namely it not only improves the flexibility of search expression, and helps the user to locate the component, but also enhances the recall and precision of component. Therefore it is significant to research the organization and retrieval mechanism of component library in the software industry.