Our publication "Performance Evolution Matrix: Visualizing Performance Variations along Software Versions" was awarded at IEEE VISSOFT 2019 in Cleveland, USA as the best conference contribution.

Programs need to run at a satisfying speed, but they get continuously changed and extended by the software developers, which can affect the performance. Our paper describes a visualization approach called "Performance Evolution Matrix", which supports software developers to find the root causes of such performance regressions. Developers can explore runtime across different versions and components of the software.

The approach was developed with collaborators from South America, from the Universidad Catóica Boliviana San Pablo, Bolivia and University of Chile. It contributes to the DFG-funded project "SoftwareDynamics²".

Details on the publication.