A visual analytics architecture for the analysis and understanding of software systems

Abstract

Visual analytics facilitates the creation of knowledge to interpret trends and relationships for better decision making. However, it has not being used widely for the understanding of software systems and the change process that takes place during their development and maintenance. This occurs despite the need of project managers and developers to analyze their systems to calculate the complexity, cohesion, direct, indirect and logical coupling, detect clones, defects and bad smells, and the comparison of individual revisions. This research considers the design of an extensible and scalable architecture to incorporate new and existing methods to retrieve source code from different versioning systems, to carry out the analysis of programs in different languages, to perform the calculation of software metrics and to present the results using visual representations, incorporated as Eclipse and Visual Studio extensions. Consequently, the aim of this work is to design a visual analytics architecture for the analysis and understanding of systems in different languages and its main contributions are the specification of the design and requirements of such architecture, taking as base the lessons learned in Maleku (A. González-Torres et al., 2016).