Streaming software analytics
Georgios Gousios, Dominik Safaric, Joost Visser · 2016
In this paper we present a novel software analytics infrastructure supporting for a combination of three requirements to serve software practitioners in utilising data-driven decision making: (1) Real-time insight: streaming software analytics unify static historical and current event-stream data enabling for immediate, nearly real-time insight into software quality, processes and users; (2) Query model: streaming software analytics substantiate for the lack of an integrated event-stream data extraction method by utilising a sophisticated, yet easy to use query Domain Specific Language; (3) Data summarisation: streaming software analytics allow for high level event-stream data summarisation in respect to distinct stakeholder groups. We expect that streaming software analytics will encourage software practitioners to move beyond information toward actionable insight, and enable for the use of analytics as a feedback and decision-support instrument, thus allowing them for an increase in quality of software systems, processes and delivery.