Dataspace Management for Large Data Sets

In an ideal case, Big Data analysis will enable us to learn relevant and interesting facts using large interconnected data sets. Dataspace support platforms and dataspace management systems have been proposed to help analysts bring together data related to the analyst’s interests. In this paper, we provide an example of such a platform. In addition to storing data and description of its characteristics, the platform supports verifying compatibility (and eventually summarizability) of the underlying data. This will help the analysts discover mistakes and prevent meaningless aggregations. As an example of utilizing the platform, we present a case of large data sets (tens of millions of observations), describe how the data sets can be used, and study the platform’s performance.