Structure-aware Fisheye Views for Efficient Large Graph Exploration.

Area of Science:

• Computer Science
• Information Visualization
• Graph Theory

Background:

• Traditional fisheye views for large graph exploration suffer from significant distortions, impairing readability of paths and structures.
• Existing methods often fail to preserve graph topology during visual exploration.

Purpose of the Study:

• To develop a novel framework for structure-aware fisheye views that minimizes distortions in large graph visualization.
• To enhance the readability of graph structures and user performance during interactive exploration.

Main Methods:

• Proposing a framework that uses edge orientations as constraints for graph layout optimization.
• Designing and implementing a family of new fisheye lenses (polyfocal, cluster, path lenses).
• Utilizing a GPU implementation for interactive processing of large graphs (up to 15,000 nodes).

Main Results:

• Demonstrated reduction in spatial and temporal distortions during fisheye zooms.
• Significant improvement in the readability of graph structures and paths.
• Achieved interactive rates for visualizing large graphs.

Conclusions:

• Structure-aware fisheye views offer a substantial improvement over traditional methods for large graph exploration.
• The proposed framework enhances user performance and layout readability, validated through evaluations and user studies.