Shortening retrieval sequences in browsing-based component retrieval using information entropy

Reuse repositories are an essential element in component-based software development (CBSD). Querying-based retrieval and browsing-based retrieval are two main retrieval mechanisms provided in real world reuse repositories, especially web-based repositories. Although browsing-based retrieval is superior to querying-based retrieval in some aspects, the tedious retrieval process is its main drawback, because the browsing-based component retrieval usually involves long retrieval sequences. In this paper, we propose a novel approach to shorten the retrieval sequences in browsing-based component retrieval using information entropy. The basic idea of our approach is to build a navigation model by ranking the features into a tree structure using the components' indexing information. According to our experimental results on real data, our approach can effectively shorten the average length of retrieval sequences. (C) 2005 Elsevier Inc. All rights reserved.Computer Science, Software EngineeringComputer Science, Theory & MethodsSCI(E)EI0ARTICLE2216-2307