Visualization for Data Scientists: How specific is it?

Abstract

Data Science has been widely used to support activities in diverse domains as Science, Health, Business, and Sports, to name just a few. Theory and practice have been evolving rapidly, and Data Scientist is currently a position much in demand in the job market. All this creates vast research opportunities, as well as the necessity to better understand how to prepare people as researchers and professionals having the background and skills to keep active in a difficult to anticipate future. While there are courses on Data and Information Visualization described in the literature, as well as recommendations by the SIGGRAPH Education Committee, they do not concern Data Science Programs and thus may not be entirely adequate to this type of Program. Besides the general concepts and methods usually addressed, a Visualization course tailored for this particular audience should probably emphasize specific techniques, tools, and examples of using Visualization in several phases along the Data Science process; moreover, it is reasonable to expect that new approaches, useful in practice, will be proposed by the Visualization research community that should be addressed in such a course. Likewise, the bibliography and teaching methods could probably be adapted. We have analyzed over forty MSc Data Science programs offered in English worldwide, and the Visualization courses most of them include, and we argue that there is a need to adapt existing recommendations and create guidelines for these courses. This panel intends to debate this topic and identify issues that need further reflection. .