A Spreadsheet Approach To Information Visualization

1
A Spreadsheet Approach To Information
Visualization

Paper by Ed Huai-Hsin Chi,Phillip Barry,John
Riedl,Joseph Konstan
2
Stages of visual Sensemaking
cycle
3
Dataset
Complexity
Volume
Numerical Spreadsheet
Visual Speadsheet
4
(No Transcript)
5

• Comparing the lifetime costs of
  hardware,software and maintenance for different
  types of computer purchases.
• Adv minimum work in visualisation

• Different views of data
• Data interactivity
• Operators like rotation
• Data filtering to a view or
• group of views
• Comparisons in data sets
• Molecular Biology
• Time series Matrices
• Algorithm Visualisation

Example
Features
Domains
6
Interactions and control flow of the spreadsheet
for Information
Visualization
7
Molecular biologists seek to determine the
interaction between gene,protein and protein
function.Similarity algorithms with visualization
provide a shortcut to find possible functions for
an unknown sequence.
8
MultiVariate Dataset
9
Time Series Matrices
Matrix 40 from PAM Series.
Matrix 120 from PAM Series
Matrix 240 from PAM Series
Matrix 62 from BLOSUM Series
10
Discovering novel patterns using multiple
visualisations representations in time-series
matrices
11
Visual Comparisons
V
Computing differences in
datasets
Looking at representations
Requires layout strategies For User Support
Difference operators at Data domain level

Menu based
Script based
12
Parallel application of direct manipulation
operators on multiple cells simultaneously
13
Ease of Comparison
Visual usage subtraction shows differences in
usage quickly
Without the Visual subtraction ,straight color
mapping does not show any difference at all.
14
Outcome of the algorithm after 5 steps
Outcome of the algorithm after 6 steps
Outcome of the algorithm after 25 steps
Outcome of the algorithm after 25 steps
Generating cells 4_1,4_2,4_3 in
the Triangulation example
15
Direct manipulation interface with menus and
dialog boxes which make system easy to use.
16
Textual commands allow users to specify
operations using simple macro languages.scripts
can be specified in a file and loaded
17
Advantages

• Simultaneous operation on all views of data.
• Eg
• Same orientation in all views.
• Highlight one point in one scatter plot and all
  corresponding points are automatically
  highlighted
• 2. Not application specific.
• 3. Suits user requirements New User vs
  Knowledgeable Viewer.