SDL 2017: Model-Driven Engineering for Future Internet

To understand more and more complex software systems and the rules that govern their development, software visualization uses more and more complex, but static visual representations (charts) to allow computer scientists to analyze complex multi-modal, multi-variant, and potentially temporal data gathered from software artifacts. Data scientist however, use interactive visual analysis to not only visualize data but to explore and understand data via interactive visualizations. In this paper, we present a language that allows us to quickly create such interactive visualizations for software. We present a process to measure software and gather data, a common data meta-model, four principal ways to combine individual charts into an interactive visualization, the language constructs needed to specify interactive visualizations, and a working implementation and examples for this language.