Analyze That! Rethinking Questions for Data Scientists in Software Engineering

In 2014, Microsoft came up with a list of 145 questions that software engineers would like data scientists to answer. Following this open call to the research community, many studies followed. Now in 2019, five years since the original study came out, no other study checked if the questions posted byMicrosoft hold for other software companies (including companies with a different focus). More so, do the problems identified five years ago still hold? Along these lines, this paper presents another list of questions that software engineers at ING would like data scientists to answer. ING is a Netherlandsbased software-defined enterprise providing banking solutions. At present, ING houses over 15,000 employees in its software division providing home-grown software solutions. As a banking company, now dealing with the complexities of software development, we replicated the original Microsoft study at ING and compared the questions. Furthermore, we contrasted the two outcomes, in light of the differences in context and time. The study shows that the two sets of data-driven software engineering questions are remarkably similar, albeit with certain key differences. CCS CONCEPTS • General and reference→ Surveys and overviews.