AutARG: Automatic Algorithms for Result Generation in Visualization

The main goal of the proposed Dissertation Fellowship project AutARG (Automatic Algorithms for Result Generation in Visualization) is to speed up the analysis process of 3D/4D simulation data from different domains like engineering, meteorology, climatology, or medical applications. Since the number and size of simulations in these fields is increasing rapidly fewer and fewer time can be spent in the actual analysis process of the generated data. We therefore want to introduce novel approaches to semi-automatically generate visualization and analysis results based on user-defined templates, rules and semantic information gathered during previous interactive analysis sessions. This will greatly improve the efficiency of the post-processing phase necessary to document and present information gained from the simulation results. In the course of this project we plan to research efficient and user friendly methods for the analysis and presentation of simulation results, which will be implemented in the SimVis framework. SimVis is a state of the art interactive visual analysis tool for large and complex simulation data, which up to now, like many other state of the art systems, is relying mostly on time-consuming user interaction. The planned project will be structured into three larger phases of research and development. In the first phase automatic generation of visualization results based on manually specified features will be introduced. The second phase will deal with semi-automatic feature specification based on templates additionally to the automatic result generation. The third phase should extend the proposed methods to incorporate story telling functionality which can be used to present results to non-experts. The results of this project will not only include the PhD thesis of Philipp Muigg and a larger number of high-quality publications, but also newly developed software modules, which shall be capable of demonstrating the full potential of the newly developed methods also for real world datasets from various application fields.