Track 1: Software Analytics

The research area of software analytics seeks to leverage data collected from software engineering processes to improve the effectiveness and efficiency of these processes. Data collected for these purposes include issues, log data, source code repositories, epic descriptions, etc. Thanks to the abundance of data, it becomes increasingly viable to apply machine learning techniques (e.g., random forests, support vector machines, neural networks) to use historic development information to support current development activities.

In the ING context, this is particularly relevant for the over 600 squads involved in software development. A key concern is epic predictability, which will be addressed within this track. Another area of interest is dependency management, between squads, software libraries used, and in the context of larger software ecosystems.

The research methods used in this track will include statistical analysis of development (CDaaS) data, enriched with survey data collected from development teams. Outcomes will include prototype tools as well as guidelines on how to make the software development process itself data-driven.

Related work:

  • Elvan Kula, Ayushi Rastogi, Hennie Huijgens, Arie van Deursen, Georgios Gousios: Releasing fast and slow: an exploratory case study at ING. ESEC/SIGSOFT FSE 2019: 785-795

  • Hennie Huijgens, Ayushi Rastogi, Ernst Mulders, Georgios Gousios and Arie van Deursen. Analyze That! Rethinking Questions for Data Scientists in Software Engineering. Technical Report TUD-SERG-2019-003. Delft University of Technology, 2019.

  • Hennie Huijgens, Davide Spadini, Dick Stevens, Niels Visser, Arie van Deursen: Software analytics in continuous delivery: a case study on success factors. ESEM 2018: 25:1-25:10

Ph.D. Student: Elvan Kula

Track leader: Georgios Gousios