Edzer Pebesma (Münster), Christian Freksa (Bremen)
Availability of more, larger and more heterogeneous systems of real-time sensors makes finding patterns in the sensed data and setting up correct response to these patters more and more complicated. Designing operational real-time systems that can cope with this involves both the actual detection of patterns in the space-time data, the semantic description of the processes sensed and patterns detected, as well as a responsible way of dealing with mis-detections, uncertainties, approximation errors, and so on. Appropriate ways of visualisation high-dimensional data, along with the patterns detected and model residuals forms integral part of this.
Research challenges address the design, prototyping and mathematical and algorithmic development for full operational information systems that sensibly extract useful information (statistics, messages, patterns, graphs, visual representations) from large real-time spatio-temporal information flows.
Possible PhD Topics
- real-time statistical analysis and visualisation of complex spatio-temporal environmental data sets
- conceptual and semantic integration of spatio-temporal data analysis systems with decision making processes
- semantic mining of complex spatio-temporal environmental data sets.