What is Data Science?

Data Science, a new discovery paradigm, is potentially one of the most significant advances of the early 21st century. Originating in scientific discovery, it is being applied to every human endeavor for which there is adequate data. While remarkable successes have been achieved, even greater claims have been made. Benefits, challenge, and risks abound. The science underlying data science has yet to emerge. Maturity is more than a decade away. This claim is based firstly on observing the centuries-long developments of its predecessor paradigms - empirical, theoretical, and Jim Gray's Fourth Paradigm of Scientific Discovery (Hey, Tansley & Tolle, 2009) (aka eScience, data-intensive, computational, procedural); and secondly on my studies of over 150 data science use cases, several data science-based startups, and, on my scientific advisory role for Insight1 , a Data Science Research Institute (DSRI) that requires that I understand the opportunities, state of the art, and research challenges for the emerging discipline of data science. This chapter addresses essential questions for a DSRI: What is data science? and What is worldclass data science research? A companion chapter On Developing Data Science (Brodie, 2018b) addresses the development of data science applications and of the data science discipline itself