CS 725/825 - Information Visualization
Spring 2016: Wednesdays, 9:30am-12:15pm, E&CS 2120

Print - Admin

Home

Staff

Blackboard

Syllabus

Schedule

Paper Presentations

Project
  Demo Schedule

Links


Tableau's data visualization software is provided through the Tableau for Teaching program.

Course Overview

The main goal of this course is to equip you with the background and tools needed to develop effective visualizations in your own research and future work. Part of developing effective visualizations requires analyzing existing visualizations and visualization problems.

One important piece of developing an effective visualization is knowing what not to do. In addition to studying recommended approaches, this course should also prepare you to rule out visualization approaches where there are mismatches in human capabilities or perception or mismatches with the intended task.

Course Organization

This course will be organized based on the "flipped classroom" model. Students will be assigned readings and homework that will be due before class time. There will be few, if any, lectures by the instructor. Class time will be spent on discussions of the assignments, student presentations, and in-class assignments. It is essential that each student be prepared to fully participate in class discussions each week.

For more details on specific assignment types and grading policies, see the syllabus.

Announcements, submission of assignments, outside-class discussion, and grading will be done on the class Blackboard site.

The required textbook for this course is Visualization Analysis and Design by Tamara Munzner.

Online Sections

There are some students participating in the course online. The audio and projected slides/images from our class meetings will be recorded using WebEx. Online students may join the WebEx session during class time or they may listen to the session afterwards (links will be posted in Blackboard). All students will participate in the same online discussions using Blackboard. Deadlines are the same for in-class and online students. If something is due before the next class meeting, that means it is due before 9:30am Eastern on Wednesday.

If you are enrolled in the on-campus sections, you are expected to attend and participate in class each week. In-class participation is part of your grade.

Course Objectives

After completing this course, you should be able to do the following:

  • Explain at a high-level the "why-what-how" framework for analyzing visualization use.
  • Given a visualization, identify the actions the vis allows and the targets of those actions.
  • Transform domain-specific tasks into the task abstraction framework.
  • Given an academic paper, identify the presence or absence of the four levels of vis design and validation.
  • Describe the channels of visual encoding and order them from most effective to least effective.
  • Identify a visualization where an inappropriate arrange design choice was made and explain why the choice was inappropriate.
  • Explain the importance choosing an appropriate colormap.
  • Use R, D3, and Tableau to create simple visualizations.
  • Use D3, Tableau, or another toolkit to create an effective interactive web-based visualization of real-world data.

General Schedule and Topic Objectives

The goal is to cover one chapter per week. The general schedule along with objectives for each chapter is given below.

Week 1 (Jan 13) - Introduction - Ch 1

  • Define visualization.
  • Explain the importance of humans in the visualization process.
  • Explain why human vision is particularly well-suited for information transfer.
  • Give an example of a visualization idiom.
  • Explain why it is best to consider multiple alternatives for vis before selecting a solution.
  • Explain at a high-level the "why-what-how" framework for analyzing visualization use.
  • Describe at least one historical visualization and explain its impact.
  • Differentiate between R, D3, and Tableau and describe the type of tasks for which each tool might be most appropriate.

Week 2 (Jan 20) - Data - Ch 2

  • Distinguish among the four basic dataset types.
  • Distinguish among the five core data types.
  • Distinguish between categorical and ordered attributes.
  • Distinguish between ordinal and quantitative attributes.
  • Explain why understanding the dataset and data types and semantics matter for designing effective visualizations.
  • Distinguish between scientific vis and information vis in terms of how spatial data is used.
  • Explain the difference between a flat table and a multidimensional table.
  • Explain some of the complexities of dealing with temporal data.
  • Identify two tools for cleaning data.

Hands-on with Tableau, R, D3

  • Use Tableau to create different views of a dataset for exploration
  • Use R to create a scatterplot matrix of a dataset suitable for examining the relationships between multiple variables
  • Use D3 to create an interactive view of a dataset
  • Explore the different types of graphs that can be created with R and D3

Week 3 (Jan 27) - Tasks - Ch 3

  • Discuss the strengths and limitations of vis tools that are for a specific purpose and those that are general.
  • Distinguish among the three levels of actions in the task abstraction framework.
  • Given a visualization, identify the actions the vis allows and the targets of those actions.
  • Transform domain-specific tasks into the task abstraction framework.

Week 4 (Feb 3) - Analysis - Ch 4

  • Explain the importance of validation in vis design.
  • Explain how the four levels of vis design fits into the what-why-how model.
  • Explain each of the four levels of vis design.
  • Distinguish between the top-down and bottom-up approaches to vis design.
  • Describe how the analysis framework applies to both top-down and bottom-up approaches.
  • Describe the four classes of threats to validity.
  • Describe validation approaches at each of the four levels.
  • Given an academic paper, identify the presence or absence of the four levels of vis design and validation.

Week 5 (Feb 10) - Marks and Channels - Ch 5

  • Explain how marks and channels are related.
  • Distinguish between the identity channel type and the magnitude channel type and indicate which channels belong to each type.
  • Distinguish between the principles of expressiveness and effectiveness in visual encoding.
  • List the channels for ordered attributes in order from most effective to least effective.
  • List the channels for categorical attributes in order from most effective to least effective.
  • Describe the effects of accuracy, discriminability, separability, popout (preattentive processing), and grouping and give one example that illustrates each.
  • Explain the implication of Stevens' Law for visualizations.
  • Explain the implication of Weber's Law for visualizations.

Week 6 (Feb 17) - Rules of Thumb - Ch 6

  • Explain potential difficulties with the use of 3D visualization.
  • Identify situations in which the use of 3D visualization would be appropriate.
  • Explain why "eyes beat memory".
  • Explain what happens when people experience cognitive load.
  • Define change blindness.
  • Explain the tradeoff between resolution and immersion.
  • Explain the Shneiderman mantra "overview first, zoom and filter, details on demand".
  • Explain the alternate concept of "search, show context, expand on demand" and identify in what situations it may be more appropriate than the Shneiderman mantra.
  • Explain the importance of the design slogan "get it right in black and white".

Week 7 (Feb 24) - Arrange Tables - Ch 7

  • Explain why the arrange design choice is the most crucial visual encoding choice.
  • Explain how the concepts of express, separate, order, and align all relate to arranging tabular data.
  • For each idiom example in the text (from scatterplot to normalized stacked bar chart), identify the "what: data" properties of the idiom.
  • For each idiom example in the text, identify the "how: encode" properties of the idiom.
  • For each idiom example in the text, identify the "why: task" properties of the idiom.
  • For each idiom example in the text, identify the "scale" properties of the idiom.
  • Differentiate between line charts and bar charts and explain when each is appropriate
  • Explain some of the disadvantages of pie charts.
  • Explain how a radial layout maps to a rectilinear layout.
  • Given a particular dataset and task, suggest an idiom and explain why it might be appropriate
  • Identify a visualization where an inappropriate arrange design choice was made and explain why the choice was inappropriate.

Week 8 (Mar 2) - Maps, Arrange Networks and Trees - Ch 8.1-8.3, Ch 9

  • Describe how the arrange design choice is different with spatial data as opposed to tabular data.
  • Describe a choropleth map.
  • Identify the two main families of visual encoding idioms for arranging network data in space.
  • Describe a spline radial layout and how it differs from a node-link layout.
  • Identify the tasks for which node-link diagrams are most appropriate.
  • Describe the adjacency matrix view of a network and contrast it with a node-link view.
  • Describe a treemap.

Spring Break (Mar 9)


Week 9 (Mar 16) - Map Color and Other Channels - Ch 10

  • Describe the components of color.
  • Describe the three main types of colormaps.
  • Explain the importance choosing an appropriate colormap.
  • Given a set of data and a task, determine an appropriate colormap.
  • Identify an inappropriate use of a colormap and suggest a more appropriate one.
  • Explain why rainbow colormaps should only be used with great care.
  • For the visual channels other than color, identify which are magnitude and which are identity channels.

Week 10 (Mar 23) - Manipulate View - Ch 11

  • Describe why changing a view might aid in understanding a dataset.
  • Explain why order can make such an impact in understanding.
  • Describe some of the design choices that can be made with selection.
  • Explain the difference between selection and highlighting.
  • Describe the three components of navigation.
  • Explain the idea behind semantic zooming.
  • Given an interactive visualization, identify the interaction idioms used.

Week 11 (Mar 30) - Multiple Views - Ch 12

  • Explain the importance and usefulness of faceting data across multiple views.
  • Contrast the two major approaches to faceting information.
  • Describe the four major design choices for juxtaposed views.
  • Explain the concept of linked highlighting.
  • Describe the three alternatives for sharing data between two juxtaposed views.
  • Contrast the use of small multiples with a grouped bar chart.
  • Given a multiform visualization, identify the ways in which the data was split into multiple views and the design choices that were made.

Week 12 (Apr 6) - Reduce Items and Attributes - Ch 13

  • Explain the need to reduce data, both in terms of number of items and number of attributes.
  • Explain the difference between filtering and aggregation and the purposes of each.
  • Identify instances of scented widgets, as opposed to standard filtering widgets.
  • Contrast histograms with bar charts.
  • Explain the benefits of a continuous scatterplot for high-density data over a scatterplot that uses size to encode density.
  • Describe the components of a boxplot.
  • Explain the idea behind dimensionality reduction.
  • Explain why dimensionality reduction is especially useful for analyzing text document collections.
  • Describe some cautions when visualizing the results of dimensional reduction with a scatterplot.

Week 13 (Apr 13) - Embed: Focus + Context - Ch 14

  • Describe the idea behind focus+context views.
  • Describe the three main ways that focus+context can be employed.
  • Contrast the magic lens idiom with the fisheye lens idiom.
  • Describe the costs and benefits of distortion.
  • Given a focus+context visualization, identify whether it uses elision, superimposition, or distortion (or some combination).

Week 14 (Apr 20) - Case Studies - Ch 15

  • Use the analysis framework to decompose a vis approach into pieces that can be compared with other approaches.