Path Tracing in 2D, 3D, and Physicalized Networks.

Area of Science:

• Information Visualization
• Human-Computer Interaction
• Computer Graphics

Background:

• Traditional advice discourages 3D for abstract data like networks.
• Prior research (Ware & Mitchell, 2008) indicated 3D reduces errors in network path tracing compared to 2D.
• The benefit of 3D visualization in networks remains uncertain with enhanced 2D techniques and interactive exploration.

Purpose of the Study:

• To investigate if 3D network visualization retains its path tracing advantage over improved 2D methods.
• To evaluate the impact of interactive exploration techniques on 3D versus 2D network data.
• To explore data physicalization using 3D printed networks with augmented reality.

Main Methods:

• Study 1: Compared 2D (with edge-routing) and 3D network layouts in virtual reality (VR) using handheld controllers for 34 users.
• Study 2: Compared VR 3D networks with physical 3D printouts augmented by Microsoft HoloLens for 12 users.
• Both studies focused on path tracing accuracy and user interaction within network visualizations.

Main Results:

• Study 1 showed significantly lower error rates for path tracing in 3D VR compared to 2D VR, despite 2D enhancements.
• Study 2 found no significant difference in error rates between VR 3D and physical 3D printouts.
• Users in the physical 3D condition naturally engaged in diverse manual interactions, suggesting potential for new interaction designs.

Conclusions:

• Three-dimensional network visualization maintains an advantage for path tracing accuracy, even when 2D representations are optimized.
• Virtual reality and physical 3D models offer viable, low-error environments for network data exploration.
• Physical 3D network data with augmented reality presents opportunities for novel, tactile interaction techniques.