Collecting data from software engineering process as well as from end users is a current trend in software industry. The datasets collected can be large and complex. Better understanding of the data and causes behind the data can be achieved in many different ways. Automated data mining systems are one approach for knowledge discovery but also visualizations methods are widely used for the knowledge discovery tasks.
In this research proposal use of visualizations in software process improvement and building software analytics systems are described. Expected results of the research are guidelines for software data visualization and criterions for determining the right metrics for analytics using visualizations. Expected impact is to get analytics feasible for more companies as using visualizations in the initial step to form the metrics can reduce the amount of data that needs to be collected and analyzed.
Anna-Liisa Mattila (Tampere University of Technology), Tommi Mikkonen (Supervisor, Tampere University of Technology): Visualizations for Software Analytics
Presented at PROFES 2014 Doctoral Symposium, Helsinki