In cooperation with Fraunhofer IGD Darmstadt, we are conducting basic research in the field of Visual Analysis of Large Relational, Time-Dependent and Multi-Dimensional Data. Multi-dimensional, time-dependent relational data, whether it is used to represent a social network, an IP network, or a street network, is omnipresent. In these often very large datasets each entry may contain a large number of multi-dimensional or time-dependent subentries. As one example, health status messages of network nodes deliver for every time step an array of readings, which may need to be processed and visuallized effectively so that potential patterns can be recognized by analysts.pWell-known and proven visual representations exist for network, time-dependent and multidimensional data types each. The goal of this project is to research integrated visual analytics approaches for making sense of large data sets consisting of relational, time-dependent and multivariate facets. Research questions we are dealing with regard the definition of novel visual representations integrating these data aspects. Adaptivity of the representations should support senatic zomm operations for effective user navigation in large data sets. Furthermore, based on appropriately defined similarity notions, data reduction and layout generation will be considered in this project.
