Evaluating Graph Viewport Optimization in VR

Theoretical (Analytical):

Practical (Implementation):

Literature Work:


Overview

In this project you will design and conduct a user study with at least 16 participants that assesses a viewport optimization technique for the inspection of 3D graph visualizations in virtual reality environments. The technique of interest, as well as the visualization framework with basic study support already exists, but a complete study prototype needs to be developed to conduct the study, while the framework might require slight adaptions. 

Problem Statement

One major issue with 3D visualization is the influence of perspective on the effectiveness and expressiveness of the visualization. In the past, we developed a framework that visualizes a 3D graph in virtual reality and provides an additional visualization to guide a user to a viewport that could be beneficial for inspecting the given graph visualization. Different metrics can be deployed to 'measure' how good or bad a certain viewport is to inspect the graph. For instance, one could count the overall number of edge crossings from each perspective and with that compare all viewports/perspectives to each other. Of course, the metrics used in the final prototype are more complex and take more than one aspect into consideration. However, the large degree of freedom in selecting a composition of dimensions for a metric opens a huge landscape and needs to be evaluated quanitatively. 
The main scope of this project is to assess the usefulness and potential benefits/drawbacks of the approach by conducting a quantitative user study on the novel technique. 

Tasks

  • Plan and design a quantitative user study
  • Implement a study prototype for the controlled user study and adapt the current framework
  • Conduct the user study with at least 16 participants
  • Statistically evaluate the results of the study

Requirements

  • Basic knowledge about visual analytics
  • Advanced programming skills in Java/C#
  • Previous experience in Unity highly recommended
  • Good conceptual skills
  • Useful: Git, VR-Experience, experience with user studies

Scope/Duration/Start

  • Scope: Bachelor/Master

Contact

References

  • Peter Eades, Michael E Houle, and Richard Webber. “Finding the best viewpoints for three-dimensional graph drawings”. In: International Symposium on Graph Drawing. Springer. 1997, pp. 87–98.
  •  Michael E Houle and Richard Webber. “Approximation algorithms for finding best viewpoints”. In: International Symposium on Graph Drawing. Springer. 1998, pp. 210–223.
  • Pere-Pau V´azquez, Miquel Feixas, Mateu Sbert, and Antoni Llobet. “Viewpoint entropy: a new tool for obtaining good views of molecules”. In: ACM International Conference Proceeding Series. Vol. 22. 2002, pp. 183–188.