1

Remaining class topics◦ Information visualization◦ Data mining◦ Data Governance ◦ Implementation issues◦ Data Security and Privacy

Approaches to teaching/learning◦ Readings on WebCampus◦ Case analyses◦ Participation in class exercises◦ BI Tutorials

IS482/682: What else are we learning?

2

In “business intelligence,” what do we mean by “intelligence”?

Where does “intelligence” come from? How is “intelligence” produced?

Intelligence in Business Intelligence...

3

Goal of information visualization:

Use human perceptual capabilities to gain insights into large data sets that are difficult to extract using standard query languages

Overall View of BI Information Visualization

Enhanced understandi

ngImprove

d decision making

Clean data

Complete data

Longitudinal data

Consistent data

More data

Accurate calculation

s

Flexible design

Properties

Patterns

Relationships

Comparisons

Anomalies

Trends

Amplify cognition

Increase available brain resources◦ Enhance parallel perceptual processing◦ Offload work from cognitive to perceptual system

Reduce search time Enhance recognition of patterns

◦ Encourage recognition instead of recall◦ Enhance “chunking” into appropriate memory

sizes for both processing and recall Provide focus/emphasis

◦ Highlight images with “pop-out” effect

How does information visualization enhance understanding & amplify cognition?

What do you see?

Subjective Contours

What do you see?

Proximity Matters

What do you see?

Scale Matters

Rapid parallel processing

◦ Feature Extraction: edges, orientation, color, texture, motion

◦ Relies on commonly accepted images (cultural, personal)

◦ Transitory: Uses primarily short-term memory, but can leave

impact Serial goal-directed processing

◦ Object Recognition: visual attention & memory important.

◦ Uses both short-term memory and long-term memory

◦ More emphasis on arbitrary aspects of symbols

◦ Different pathways for object recognition & visually guided

motion

Two general stages of visual processing

10

Parallel Processing• Orientation• Texture• Color• Motion• Size

a

AB

C

D

Detection• Edges• Regions• Light• 2D Patterns

Serial Processing• Object Identification• Collation• Short Term Memory 5 ± 2 = 3 to 7 Objects

Visual properties processed without significant cognition.◦ No need to focus attention; must stand out◦ Can be perceived immediately; less than 250 ms◦ May mislead viewer; may create inappropriate

and lasting significant emphasis The visual properties are:

◦ Color◦ Motion◦ Edge segmentation; primitive features◦ Orientation◦ Size

Pre-attentive processing

Pre-Attentive ProcessingHow many 3s ?

08028085080830802809850-802808567847298872ty458202094757720021789843890r455790456099272188897594797902855892594573979209

Color Pre-Attentive (Pops out)How many 3s ?

08028085080830802809850-802808567847298872ty458202094757720021789843890r455790456099272188897594797902855892594573979209

Pre-Attentive Processing - Color

Pre-Attentive Processing - Motion

Segmentation by Primitive Features

How many areas?

Pre-Attentive Processing - Orientation

Pre-Attentive Processing - Size

Combining properties (does not pop out)

Expression recognition

21

The goal of information visualization is to enhance understanding and amplify cognition.◦ Increase available brain resources◦ Reduce search time◦ Enhance recognition of patterns◦ Provide focus/emphasis

Make best use of parallel and serial processing.◦ Understand how people process images.◦ Understand the speed with which people process

images without conscious thought.

Brief recap of class on 4/4/13

How to design effective

information visualizations

Art is valued for its originality and expressiveness.

Art is valued for pushing the bounds of accepted norms and potentially expanding the definition of those norms.

Design is valued for its fitness to a particular user or task.

Design is valued for its effectiveness and use.

Design and art are different

“An affordance is a quality of an object which allows an individual to perform an action.” (Wikipedia)◦ A knob implies twisting, a string means pulling.

In design, we look at how the affordances of an object reveal how it will be used.◦ A push plate on a door means that it should be pushed,

rather than pulled, open.◦ A line showing where the average is on a graph means

that the viewer should compare the data to the average.◦ A vertical line means someone should read a table or

graph vertically. Design is “good” when the perceived affordance

(“is for”) is equivalent to the actual affordance.

Affordance

25

Task rather than the visualization User rather than the technology Information content rather than the data Message rather than the medium Accuracy rather than beauty

Focus on the...

Analyzing◦ Discover the message in the data.

Monitoring◦ Track information and look for anomalies.

Planning◦ Prepare for the future.

Communicating◦ Send a message to another person.

Primary visualization tasks for BI

The differences. Like:◦ The difference between analyzing and

communicating.◦ The difference between monitoring and

analyzing.◦ The difference between planning and monitoring.

Think what about the differences?◦ Who does it?◦ Why is the person doing it?◦ How does it happen?◦ What are the results?

Think about...

Show the data Induce to viewer to think about the data Avoid distorting what the data have to say (next 3

slides) Present many numbers in a small space Encourage the eye to compare different pieces of data Reveal the data at several levels of detail, from

overview to fine structure Make large data sets coherent Serve a clear purpose Be closely integrated with the statistical and verbal

descriptions of a data set.

What visualization should do (Tufte 1983)

Example

20022001200019991998

5000

4750

4500

Stock market crash?

Example

20022001200019991998

5000

2500

0

Show entire scale

Example

20001990198019701960

5000

2500

0

Show in context

2010

32

http://www.gapminder.org/ http://visual.ly/

http://www.babynamewizard.com/voyager http://www-958.ibm.com/software/data/co

gnos/manyeyes/

Sample information visualizations

33

Tufte’s principles (short and sweet) Nielsen’s ten usability heuristics (long and

detailed)

Heuristics galore!!

34

Provide the greatest number of ideas in the shortest time with the least ink in the smallest space

Enforce visual comparisons◦ Show comparisons adjacent in space◦ Show causality◦ Show multivariate data◦ Use direct labeling: Avoid separate legends and keys◦ Use small multiples (called trellis chart): a number of

small, simple adjacent charts to encourage comparison

Avoid “chart junk”

Tufte’s principles

1. Visibility of System Status ◦ Always keep users informed about what is going on.◦ Provide appropriate feedback within reasonable time.

2. System Matches Real World ◦ Speak the users' language, with words, phrases and

concepts familiar to the user, rather than system-oriented terms.

◦ Follow real-world conventions, making information appear in a natural and logical order.

3. User Control and Freedom ◦ Users often choose system functions by mistake.◦ Provide a clearly marked "out" to leave an unwanted state

without having to go through an extended dialogue. ◦ Support undo and redo.

Nielsen’s ten usability heuristics

4. Consistency and Standards ◦ Users should not have to wonder whether different words,

situations, or actions mean the same thing. ◦ Follow platform conventions.

5. Error Prevention ◦ Even better than good error messages is a careful design

which prevents a problem from occurring in the first place.

6. Recognition rather than Recall ◦ Make objects, actions, and options visible. ◦ User should not have to remember information from one

part of the dialogue to another. ◦ Instructions for use of the system should be visible or

easily retrievable whenever appropriate.

Nielsen (cont.)

7. Flexibility and Efficiency of Use ◦ Accelerators -- unseen by the novice user -- may often speed

up the interaction for the expert user so that the system can cater to both inexperienced and experienced users.

◦ Allow users to tailor frequent actions.

8. Aesthetic and Minimalist Design ◦ Dialogues should not contain information which is irrelevant

or rarely needed. ◦ Every extra unit of information in a dialogue competes with

the relevant units of information and diminishes their relative visibility.

9. Help users Recognize, Diagnose, and Recover from Errors ◦ Expressed in plain language (no codes)◦ Precisely indicate the problem◦ Constructively suggest a solution.

Nielsen (getting closer...)

10.Help and Documentation ◦ Even though it is better if the system can be

used without documentation, it may be necessary to provide help and documentation.

◦ Help information should be easy to search, focused on user's task, list concrete steps to be carried out, and not be too large.

Nielsen (the last!)

The purpose of the chart is to compare the sales in Company G to the rest of the companies. Does it do a good job?

Better in 3D?

The purpose is to track quarter-to-date performance of sales representatives for a sales manager, including relative performance. Does it work?

The purpose is to compare one product’s sales to other products. Good or bad visualization?

The purpose is to display sales revenue in the state of Kansas associated with 12 products across the four quarters of a year. Does it do a good job?