BarcodeTree: Scalable Comparison of Multiple Hierarchies

Area of Science:

• Computer Science
• Information Visualization
• Data Visualization

Background:

• Comparing topological structures and node attributes of multiple trees is challenging.
• Existing methods often require aggregation or lack scalability for numerous trees.

Purpose of the Study:

• To introduce BarcodeTree (BCT), a novel visualization technique for comparing multiple trees.
• To enhance the simultaneous overview and comparison of numerous shallow and stable trees.
• To facilitate the understanding of topological structures and node attribute values.

Main Methods:

• Developed BarcodeTree (BCT), a row-based visualization resembling a barcode.
• Designed visual cues and interactive techniques for tree comparison.
• Aligned matching nodes horizontally and stacked trees vertically for space efficiency.

Main Results:

• BCT allows simultaneous overview of up to one hundred trees without node aggregation.
• Experimental comparison with icicle plots indicated BCT reduces vertical distance, easing tree comparison.
• Case studies demonstrated BCT's utility with large datasets of hundreds of trees.

Conclusions:

• BarcodeTree (BCT) offers an effective and space-efficient method for visualizing and comparing multiple tree structures.
• BCT enhances visual comparison by minimizing vertical separation between trees.
• The technique proves valuable for analyzing large collections of trees in various applications.