Marti Hearst

1
SIMS 247 Lecture 3Graphing Basics, Continued

• January 27, 1998

2
Today

• Finish graphing basics
• Demonstrate on web access example
• Discuss Tuftes Data Ink Maximization principle

3
Types of Symbolic Displays(Kosslyn 89)

• Graphs
• Charts
• Maps
• Diagrams

4
Types of Symbolic Displays

• Graphs
• at least two scales required
• values associated by a symmetric paired with
  relation
• Examples scatter-plot, bar-chart, layer-graph
• Charts
• discrete relations among discrete entities
• structure relates entities to one another
• lines and relative position serve as links
• Examples family-tree, flow-chart, network diagram

5
Types of Symbolic Displays (cont.)

• Maps
• internal relations determined (in part) by the
  spatial relations of what is pictured
• labels paired with locations
• Examples map of census data, topographic maps
• Diagrams
• schematic pictures of objects or entities
• parts are symbolic (unlike photographs)
• Examples how-to illustrations, figures in a
  manual

6
Standard Graph Types

• Scatter plots
• Line graphs
• Time series (strip charts)
• Dot plots
• Bar Charts
• Pie Charts
• Layer Graphs

7
Anatomy of a Graph

• Framework
• sets the stage
• kinds of measurements, scale,
• internal and external
• Content
• point symbols, lines, areas, bars,
• Labels
• title, labels for axes, tic marks, legends,
• Background
• does not impart essential meaning

8
Choosing the Axes

• Independent vs. Dependent variables
• the dependent variable changes relative to the
  independent one
• sales against season
• tax revenue against city
• What happens when there is more than one
  independent variable?
• Most important is assigned to X axis
• Other(s) differentiated by mark symbol

9
Basic Types of Data

• Qualitative -- nominal
• no inherent order (for comparisons)
• city names, types of diseases, ...
• Qualitative -- ordinal
• ordered, but not at measurable intervals
• first, second, third,
• cold, warm, hot
• Quantitative -- interval and ratio

10
Combining Data Types in Graphs(adapted from
Kosslyn 89)
Nominal by Nominal Use a Chart
11
Scatter Plots

• Qualitatively determine if variables
• highly correlated
• linear mapping between horizonal vertical axes
• nonlinear relationship
• a curvature in the pattern of plotted points
• low correlation
• spherical, rectangular, or irregular
  distributions
• Place points of interest in context
• apply shapes or color to points representing
  special entities, see where they end up

12
Time Series

• Change over time
• Facilitates finding trends
• Also known as strip charts

13
Web Page Visit Behavior

• What are our goals?
• What questions do we want to answer?
• What kind of data might we collect?
• How might we convey this information?
• Who is the audience?

14
Web Access Data Types(consider the possible
combinations)
15
Hypothetical Graphs
of accesses
of accesses
length of access
URL
url 1 url 2 url 3 url 4 url 5 url 6 url 7
45
40
35
of accesses
30
length of access
25
20
15
10
5
0
long
very
long
short
of accesses
medium
days
length of page
16
How to Show Link Traversal?

• How to link together the to-links and from-links
  in our web access example?

17
Charts

• Structural / organizational material
• nominal by nominal
• Specify relationships among discrete members of a
  set
• Not relating on quantitative dimensions
• Components of Charts (Kosslyn 89)
• directed vs. undirected links
• how many types of links
• types of mapping
• e.g., one-to-one, one-to-many, many-to-many
• Tables can also be considered charts

18
Mapping Types in Charts
one-to-one
one-to-many
many-to-many
19
Chart Example(organizational chart)
20
Chart Example(Software architecture, labels
omitted, by Chen and Hong 97)
21
How to show link patterns in web access example?
Problem only shows one step Think about this for
next time.
22
Graph/Chart Hybrids

• An area for innovation
• Combine Structure with Graphics
• Example Docuverse (Spring et. al 96)
• structure file system structure
• graphics color -gt file age
• Example TileBars (Hearst 95)
• structure document subtopics (columns)
• structure faceted query (rows)
• graphics gray-level -gt number of hits

23
Docuverse (Spring et. al 96)
24
TileBars (Hearst 95)
25
Discussion Tuftes Notion of Data Ink
Maximization

• What is the main idea?
• draw viewers attention to the substance of the
  graphic
• the role of redundancy
• principles of editing and redesign
• Whats wrong with this? What is he really
  getting at?

26
Next TimeMultidimensional Graphing

• How do we handle cases with more than three
  variables?
• Multiple views
• Scatterplot matrix
• Parallel Coordinates
• Tufte examples combine space and time
• Interaction/animation across time

27
References for this Lecture

• Kosslyn, Stephen M. Understanding Charts and
  Graphs. Applied Cognitive Psychology, 3, 185-226.
  1989
• Spring, Michael B., Morse, Emile, and Heo,
  Misook. Multi-level Navigation of a Document
  Space. http//www.lis.pitt.edu/spring/mlnds/nlnd
  s/mlnds.html
• Schall, Matthew. SPSS DIAMOND a visual
  exploratory data analysis tool. Perspective, 18
  (2), 1995. http//www.spss.com/cool/papers/diamon
  dw.html
• Hearst, M. TileBars, Visualization of Term
  Distirubtion in Full Text Information Access.
  Proceedings of ACM SIGCHI 95.
• Tufte 83