Multifaceted Data Visualization: A unifying concept for analyzing spatio-temporal data ensembles

• Applicant: Professor Dr.-Ing. Lars Linsen
• Subject Area: Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 545837520

A main objective of scientific research is to create models that describe natural phenomena. Typically, such models introduce parameters, where the parameter values are not a priori known. Instead, simulations are run with multiple parameter settings for the given model leading to a simulation ensemble. Then, it is a data analysis task to interpret the simulation outcome. An interactive, user-centric approach allows the scientist to use domain knowledge to steer the analysis and interpret the findings. Visualization methods facilitate such an analysis. Given that natural phenomena are typically spatio-temporal, the ensemble consists of multifaceted data, where the facets refer to space, time, and parameters. Commonly, analysis methods are developed that target a subset of the facets. In this project, we propose a unifying concept for analyzing the multifaceted data of spatio-temporal ensembles. We propose a common and scalable data management in the form of what we call the multifaceted hypercube. We further propose different interactive visual analysis methods for the multifaceted hypercube. They target overview visualizations, direct visualizations of all facets and flexibly adjustable combinations thereof, interactive and automatic feature extraction from the multifaceted hypercube, uncertainty-aware visualizations of multifaceted regions as well as comparative visualizations of multiple multifaceted hypercubes. Thus, we are proposing a holistic approach for the main analysis objectives of simulation ensembles that require a joint consideration of all facets. We plan to evaluate our approach within real-world scenarios from different domains.

• DFG Programme: Research Grants