Award winning work in Software Metrics
Software metrics is measuring and evaluating different properties of a software to, for example, speed up product development. In the next step, researchers want to apply machine learning and artificial intelligence to the results to maximize the insight provided by software metrics.
Software metrics is measuring and evaluating different properties of a software to, for example, speed up product development. In the next step, researchers want to apply machine learning and artificial intelligence to the results to maximize the insight provided by software metrics.
Miroslaw Staron, Professor at the Department of Computer Science and Engineering, works with colleagues at Ericsson in Gothenburg and the Poznan University of Technology, Poland, to establish, maintain and evolve measurement programs in large software development companies. The collaboration was established in 2006 and has resulted in over 40,000 measurement systems and dashboards used by industry.
"For example, we have implemented measurements on release readiness to motivate organizations to break old habits of 'plan release' and move on to 'delivering as fast as possible when the customer wants the latest software' and thus become more agile and proactive in development. Another example is our work on using machine learning to identify signal malfunctions in mobile networks and then prioritize the measures" he says.
Now, the collaboration with Ericsson and the Poznan University of Technology has been awarded the Team Impact Commitment award in the first ever Real Impact Awards, an international prize founded by Emerald Publishing, with the aim to raise the profile of individuals, teams and institutions that have placed real impact at the top of their agenda; recognize innovative approaches to impact; celebrate interdisciplinary research; bridge the gap between research and practice; and tell real impact success stories and showcase those driving the debate. "For us, the price is a confirmation that our way of using industry-related problems for theory-building works and is important, and we are sure that the award can provide more visibility to academics and practitioners working in the same way within software engineering," says Miroslaw Staron.
The motivation states, among other things, that the group have demonstrated the ability to build on the skills of academics and practitioners. In addition to the implemented systems, the team have developed courses for students and practitioners and co-authored about 40 papers and one textbook. Their current focus is on applying artificial intelligence and machine learning for maximizing the insight provided by software metrics.
"Our ambition is that autonomous, machine-learning measurement systems should be able to take over the daily work of discovering new measurements, patterns and analyzes. The focus of software engineers should be on finding new business opportunities, values and visualizations. Our new concept of AI-based autonomous measurement systems has shown great potential in areas such as source code analysis, requirement analysis and testing" says Miroslaw Staron.
Members of the winning team:
- Miroslaw Staron, University of Gothenburg
- Miroslaw Ochodek, Poznan University of Technology, Polen
- Wilhelm Meding, Ericsson
- Martin Sjödin, Ericsson