VisBig

Big Data, and especially Streaming Big Data (SBD) is often analysed by automated processes, not amenable to real-time decisions by human analysis. There, timely decision making must involve the quick and easy understanding of the data. Making that possible is the purview of Visual Analytics, but existing solutions are tailored for mostly static (comparatively) small datasets and will not scale well for SBD.

In VisBig, we will fill that gap by researching solutions that separate the wheat from the chaff and highlight real-time relevant changes in the data streams. We will develop novel visualization techniques capable of displaying important features in SBD. These will be supported by machine-learning for the identification of relevant changes in the streams, to be highlighted in the visualizations. Underlying it we will study how best to cope with SBD in terms of collection, filtering and (where necessary) storage, in ways that make the other components feasible.