Tree-Maps: Visualizing Hierarchical Data

1
Tree-Maps Visualizing Hierarchical Data

• Hench Qian Bill Shapiro
• September 21, 1999

2
Problem

• Representing hierarchical data
• Structural information
• Content information
• Objectives
• Efficient Space Utilization
• Interactivity
• Comprehension
• Esthetics

3
Favorite Sentence

• Most people come to understand the content and
  organization of these structures (hierarchical
  information) easily if they are small, but have
  great difficulty if the structures are large.

4
Traditional Methods

• Outlines (e.g., Windows Explorer)

5
Traditional Methods

• Tree Diagrams (WebView, University of Pittsburg)
• Text-based tools (e.g., ls in UNIX)

6
An Alternative Tree-Maps

• Problems with traditional methods
• Lots of dead space
• Difficult to represent large number of items
• Tree-Maps
• Utilize 100 of display space
• Scalable to thousands of items

7
Tree-Maps
8
How It Works
9
Example TennisViewer
10
Critique

• Strengths
• Utilizes 100 of display space
• Shows relationships between nodes through size
  and color
• Can effectively display large data sets
• Weaknesses
• Size comparison may be difficult
• Cluttered display
• Difficult to discern boundaries
• Shows only leaf content information

11
Other Hierarchical Visualization Tools

• Docuverse (University of Pittsburg)
• Containers (file directories) are shown as pie
  segments
• Documents (files) are shown as colored dots
  within pie segments
• Can represent a large number of items (3000)
• Utilizes a fraction of the display space

12
Docuverse
-shows 3502 files in 164 directories -pie
segments represent of links to each
directory/file -Color represents age of files
13
Other Hierarchical Visualization Tools

• Hyperbolic Trees (Xerox/Inxight)
• Dynamically warps display to zoom in on areas of
  interest
• Good for interactive browsing
• Does not provide a good overview of the entire
  dataset
• Demo