Interactive visualization of quality variability at run-time: Master's thesis

Smart environments are dynamic in nature, and the software running in these environments requires quality adaptations in order to function efficiently. The result of these adaptations, i.e., quality variability, must be verified in some way, and visualization can be used to aid this verification process. The research problem in this work was to find suitable visualization techniques to visualize quality variability and implement a visualization tool that encompasses these techniques and provides an interactive visualization of quality variability for the user. As a solution to the research problem, this work presents an interactive quality visualization tool. The requirements specification for the implemented tool was derived from the literature review and the intended usage context of the tool, i.e., smart environments. The literature review explores a set of applicable visualization techniques and compares existing visualization tools with regard to the features required to represent quality variability visually at run-time. The visualization techniques selected for the tool include interactive timelines, charts and meters that enable analysis of the quality attributes and their variability in different time ranges or points in time. Some additional visualization techniques were also included such as treemaps and graphs to visualize the structure of the smart environment. The visualization techniques include open source visualization techniques and self-made techniques designed and implemented from the start to cover the specific requirements set for the tool. The main contribution of this work is the visualization tool that can be used to visualize different quality attributes and their variability. Moreover, the tool can easily be deployed in different environments due to its architecture and the selected implementation technologies that make the solution extensible and portable. The implemented visualization tool was evaluated in the context of a smart environment in which security was adapted at run-time. The case study demonstrated that the implemented tool can be used in the analysis of the variability of different quality attributes. The trend of a single quality attribute can be studied for different time ranges or points in time according to need. The relationships between different quality attributes can also be studied with the help of appropriate visualization techniques. In addition, the visualization tool was successfully tested on mobile devices.