This book addresses the economics of Big Data, beyond a project by project cost analysis.The fuel or input of any Big Data system is source data. In many cases of current use of Big Data, excepting Open Data, the data is obtained for one single purpose, and used by a single organisation, because of the likely absence of an open and transparent market where such data can be purchased or sold. Our objective is to analyse the objects, agents, and mechanisms at play for source data within the Big Data context, and give an economic perspective rather than an Information Technology engineering perspective to Big Data.The nature of data will be discussed, as a good which is digital rather than material, can be replicated at virtually no cost, and is not burnt or consumed in any irreversible way by its being used in a computation.Requirements for exchange mechanisms of source data, and associated rights of use, will be discussed, with rules for transfer and use, control retention or not by originating owner, privacy and other basic requirements and constraints on data access. Pricing and contract requirements for categories of data rights will be described, and illustrated on a few examples.
My initial impression is very positive... Very good reflections on the new market that Big Data brings about and a qualitative and informed analysis of the different agents and mechanisms that constitute the new data market. However, I have the feeling that in order to fully achieve the objectives of the book we need much more than just qualitative information on how Big Data is used and the sort of results that are obtained. In order to build more accurate economic models of Big Data, we need, non paradoxically, much more information and data on the quantitative use of that information. Models have to be sustained by quantitative information.In addition, I think, the processing algorithms and the objectives of the data treatment become essential since they are a part of the final product.I think, therefore, it is really difficult, at least for now, to separate "big Data" from the processing algorithms and, therefore, to separate that from Information Technology. Perhaps in the medium term future, when those algorithms are fully established, this objective would be possible. However, this is just my initial reflections. I hope to be able to better understand the subject in the coming weeks.