Implementing Improved Classification and Multiple Search Criteria in Software Repository

Abstract

Every software component is prepared for some specific purpose but there may be some specific features in it, which can be reused for some other purpose as well. The problem that is generally encountered by much of the research and development in information retrieval is aimed at improving the effectiveness and efficiency of retrieval. Repositories should be designed to meet the growing and changing needs of the software development organizations as it is used for retrieving large numbers of software components. Storage and representation of reusable software components in software repositories to assist retrieval is a key concern area. Effective component retrieval from a large repository is one such aim of the ongoing research. In software component repository thousand of components are stored using various classification techniques. In this thesis various assets of the component repository like component searching mechanisms and classifications such as Free Text, Enumerated, Attribute Value, and Faceted classifications are discussed. A New classification scheme is proposed which covers the limitations of existing classification schemes. In order to do so, the first requirement is a user friendly interface which provides the best solution for reorganization of the components in the repository so that recall and precision may be maximized. A component repository system is built in which the users can perform search based on multiple search criteria. The basic comparison of search presented in this thesis on the basis of precision and recall deals with evaluating the technique for retrieval of the software components for the purpose of reuse.