The analysis of the evolution of large software systems is challenging for many reasons, such as the retrieval and processing of historical information and the large quantity of data that must be dealt with. While recent research advances have led to solutions to these problems, a central question remains: How do we deal with this information in a methodical way and where do we start with our analysis? We present a methodology based on interactive visualizations which support the reconstruction of the evolution of software systems. We propose several visualizations which help us to perform software evolution analysis of a system “in the large ” and “in the small”, and apply them to 2 large systems. KEY WORDS: Software Evolution Analysis, Software Visualization

On porting software visualization tools to

Research on Mining Software Repositories (MSR) has yielded fruitful results in many Software Engineering areas including software change comprehension, bug prediction, and developer network recovery. When performing MSR research, the first task is to extract features corresponding to source code details from repositories. Since reusable feature extraction tools are not available, each MSR research group builds their own extraction tool, a duplication of effort. We introduce a reusable feature extractor, Kenyonweb, for MSR research. Kenyon-web is fully reconfigurable, pluggable, and serves most MSR related tasks. In this report, we show the architecture of Kenyonweb and demonstrate its utility by showcasing a sample MSR task.

More than 90 % of the cost of software is due to maintenance and evolution. Understanding the evolution of large software systems is a complex problem, which requires the use of various techniques and the support of tools. Several software evolution approaches put the emphasis on structural entities such as packages, classes and structural relationships. However, software evolution is not only about the history of software artifacts, but it also includes other types of data such as problem reports, mailing list archives etc. We propose an approach which focuses on historical dependencies and defects. We claim that they play an important role in software evolution and they are complementary to techniques based on structural information. We use historical dependencies and defect information to learn about a software system and detect potential problems in the source code. Moreover, based on design flaws detected in the source code, we predict the location of future bugs to focus maintenance activities on the buggy parts of the system. We validated our defect prediction by comparing it with the actual defects reported in the bug tracking system. 1

Abstract not found