CRAFT A Framework for Evaluating Software Clustering Results in the Absence of Benchmark De.pdf

CRAFT: A Framework for Evaluating Software Clustering Results in the Absence of Benchmark Decompositions Brian S. Mitchell and Spiros Mancoridis Department of Mathematics Computer Science Drexel University, Philadelphia, PA, USA {bmitchel, smancori}@ Abstract Software clustering algorithms are used to create high-level views of a system’s structure using source code-level artifacts. Software clustering is an active area of research that has produced many clustering al- gorithms. However, we have seen very little work that investigates how the results of these algorithms can be evaluated objectively in the absence of a benchmark de- composition, or without the active participation of the original designers of the system. Ideally, for a given system, an agreed upon refer- ence (benchmark) decomposition of the system’s struc- ture would exist, allowing the results of various clus- tering algorithms to be compared against it. Since such benchmarks seldom exist, we seek alternative methods to gain confidence in the quality of results produced by software clustering algorithms. In this paper we present a tool that supports the eval- uation of software clustering results in the absence of a benchmark decomposition. 1. Introduction Background Maintaining large and complex software systems is a daunting task. Poor software maintenance practices often lead to a system’s premature retirement, or ne- cessitates significant restructuring in order to keep the source code in manageable condition. There are many causes for these poor maintenance practices. Little can be done about practical issues such as limited main- tenance budgets, lack of access to the original devel- opers and designers of the system, and short devel- opment schedules. However, if software maintainers have a thorough understanding of the underlying sys- tem structure, it is more likely that changes to the source code can be performed without deviating from the system design and architecture. Additionally, i