Exercises & Project for Information Visualization

Short Paper 1 | Due Week 4
Visualization 1 | Due Week 6 | Revision Due Week 9
Visualization 2 | Due Week 9 | Revision Due Week 12
Short Paper 2 | Due Week 10
Heuristic Evaluation | Due Week 12
DataStory Paper | Due Week 13

Project | Due Week 15

Student Work: Class Exercises

Short Paper 1 – Data Journalism Handbook (7.5% of Course Grade)

Read listed article in these chapters in the Data Journalism Handbook v2: https://datajournalismhandbook.org/handbook/two
- Doing Issues with Data: From Coffee to Colonialism: Data Investigations into How the Poor Feed the Rich
- Assembling Data: Reassembling Public Data in Cuba: How Journalists, Researchers and Students Collaborate When Information Is Missing, Outdated or Scarce
- Working with Data: Accounting for Methods in Data Journalism: Spreadsheets, Scripts and Programming Notebooks
- Experiencing Data: Data Visualisations: Newsroom Trends and Everyday Engagements
- Investigating Data: Telling Stories with the Social Web
- Organizing Data Journalism: Archiving Data Journalism
- Situating Data Journalism: Data Journalism with Impact
1. Write Short Report (3-4 page, single spaced, 10pt font) using this handbook report template and concisely answer these questions:
Key Insights: What are the key take–aways for you?
What could be useful for you in the future (such as course assignments) that you learned each assigned article?
What surprised you when reading the assigned chapters of the Data Journalism Handbook?
What was not clear? Which concepts need further explanation?
Submit Report via the dropbox in Week 4 by Sun Feb 16 the latest.

Visualization 1 – Create Motion Charts using Tableau (12.5% of Course Grade)

Get Data
Week 5 - Resources: Public Data
Format Data so that it can be visualized.
Preparing Excel Files for Analysis using Tableau
How to Prepare Data for Motion Chart (9:52)
Visualize Data using Motion Chart
Motion Chart using Tableau
Motion Chart
Ex1 = Vis 1 - Gapminder / Google Motion Chart (15:35) Useful intro but not using Google's Motion Chart in Ex1
How to Prepare Data for Motion Chart (9:52)
Tableau
Creating Motion Chart using Tableau (15:12)
Tutorials in Resources in Week 5 on Lectures page: http://comminfo.rutgers.edu/~aspoerri/Teaching/InfoVisOnline/Lectures/Lectures_noTabs.htm#tableau

Requirements
Need to create at least two Motion Charts (can be side-by-side or on two different worksheets) in Tableau.
Final Data Set needs to satisfy these requirements:
Entity dimension: country or city or ... at least 25 different entities
Time dimension: at least 10 years and/or 12 months. The greater the time period, the better.
At least three independent data dimensions that can be displayed in a scatterplot and that change over time and you believe that there can be a possible correlation between them.
At least one categorical data dimension.
At least one relevant data dimension than be used for size coding (e.g. population of a country) and this data dimension does not have to change over time, but it can.
In Total, at least six data variables.
Report:
- Create a Google Docs document (or create Google Site or PBworks site) and describe:
  1. Why you created the data set; why you selected the data variables and how they help you address the question you want to address.
  2. Your process of finding and preparing the data;
  3. Insights you gained by interacting with the data using Motion Chart (e.g. relationships?, correlations?, trends?, outliers?) and include screenshots.
- Upload Tableau .twbx file (File > Export Packaged Workbook) to Google Drive and add link to report
  (make sure to set permissions so that anybody with URL can access it).
Hand-in: URL for Report (make sure to set permissions so that anybody with URL can access it)
in Week 6 by Sun Mar 1 the latest.
(Revision is due in Week 9 by Mar 29 the latest).

Visualization 2 – Design Dashboards using Tableau (12.5% of Course Grade)

Get Data
Week 5 - Resources: Public Data
You can build on data used for Visualization 1 if the data contained enough relevant independent data variables.
If you can assemble a new data set with vaiables related to a different topic is you like.
Format Data so that it can be visualized
Preparing Excel Files for Analysis using Tableau
How to Prepare Data for Motion Chart (9:52)
Visualize Data and Create Dashboards using Tableau
Overview (4:05)
Tableau
Tutorials in Resources in Week 5 on Lectures page: http://comminfo.rutgers.edu/~aspoerri/Teaching/InfoVisOnline/Lectures/Lectures_noTabs.htm#tableau
Requirements
Need to create
at least three Motion Charts (can be side-by-side or on different worksheets in Tableau)
and at least three Bar / Line / Treemap and/or Map Charts
and at least two Dashboards that incorporate the created displays in an effective way
Visual DataStory Principles (18:58)
and makes sure the displays are linked by Highlighting:
http://onlinehelp.tableau.com/current/pro/desktop/en-us/help.htm#actions_highlight.html
Final Data Set needs to satisfy these requirements:
Entity dimension: country or city or ... at least 25 different entities
Time dimension: at least 10 years and/or 12 months. The greater the time period, the better.
At least five independent data dimensions that can be displayed in a scatterplot and that change over time and you believe that there can be a possible correlation between them.
At least one categorical data dimension.
At least one relevant data dimension than be used for size coding (e.g. population of a country) and this data dimension does not have to change over time, but it can.
In Total, at least eight data variables.

Report:
- Create a Google Docs document (or create Google Site or PBworks site) and describe:
  1. Why you created the data set; why you selected the data variables and how they help you address the question you want to address.
  2. Your process of finding and preparing the data;
  3. Insights you gained by interacting with the data using Dashboards (e.g. relationships, correlations, trends, outliers?) and include screenshots.
- Upload Tableau .twbx file (File > Export Packaged Workbook) to Google Drive and add link to report
  (make sure to set permissions so that anybody with URL can access it).
Hand-in: URL for Report (make sure to set permissions so that anybody with URL can access it)
in Week 9 by Sun Mar 29 the latest.
(Revision is due in Week 12 by Sun Apr 19 the latest).

Short Paper 2 – Search Interfaces OR Machine Learning Visualizations (7.5% of Course Grade)

OPTION 1 Read the chapter Information Visualization for Search Interfaces (2009) provides excellent historical overview of the different types of visual search interfaces that have been proposed.
OPTION 2 Read at least TWO of these papers related to Machine Learning Visualizations:
Embedding Projector: Interactive Visualization and Interpretation of Embeddings
How to Use t-SNE Effectively
Visualizing Dataflow Graphs of Deep Learning Models in TensorFlow
Write Short Report using this short paper 2 template and concisely answer its questions.
Submit Report via the dropbox in Week 10 by Sun Apr 5 the latest.

Heuristic Evaluation of InfoVis Tool (15% of Course Grade)

You will conduct a Heuristic Evaluation of a tool or website TBA using Nielsen's 10 Heuristics.
A template will be provided that you will use to write a short report (4-5 pages) that identifies usability issues and suggests improvements.
Specific instructions to be provided shortly.
Submit Report via the dropbox in Week 12 by Fri Apr 17 the latest.

DataStory Paper (7.5% of Course Grade)

You will write a paper that describes your understanding, based on the materials covered and the topics discussed in class, how to design an effective “DataStory” and provide examples.
The goal is to combine and synthesize the papers by all students to hopefully arrive at a “DataStory Manifesto”.
Submit Report via the dropbox in Week 13 by Sun Apr 26 the latest.

Project (25% of Course Grade)

You can choose between several Individual Project options or a Group Project.

Screencast: Brief Overview of Possible Term Projects

Options
Individual Project options
Review and Analyze existing visualization tools for a specific data domain.
Create a How to Guide that describes how to visualize certain data sets and which tools can be used.
Evaluate InfoVis tool: conduct a user study to evaluate a visualization tool.
Create InfoVis Prototype: creating or enhancing a visualization tool.

(Group) Data Story Project
Collect Large Data Set for specific data domain and/or Use Data APIs to access large data domain.
Use Visualization Tool(s), such as ManyEyes, Google Motion Charts, Tableau, and/or Google Fusion Tables to visualize data.
Develop DataStory for specific data domain and presented data.

Timetable
Proposal Deadline = Week 7. Send instructor 1 page proposal.
Term Project Report Deadline = Week 15 = May 6

Grading
Quality of actual project (60%)
Class presentation (20%)
Quality of write-up of results (20%)

Project Descriptions

Review & Analyze
What to Do
Specify Data Domain to be visualized.
Define Tasks to be supported by information visualization tools.
Research state-of-art approaches. Identify 3 to 5 distinct "clusters" of tools to highlight & analyze.
Explain which features you used to distinguish between the different "clusters."
Select the "best" or "most representative" tool for each "cluster."
Provide Analysis and apply "InfoVis Toolbox" for each selected tool.
Think of yourself as "information visualization tools curator" and provide your personal analysis and "point of view" of what matters.
Create Presentation
Create set of slides to showcase the 3 to 5 distinct approaches you have identified.
Provide visual examples of the tools presented.
Create Report
20 to 25 pages, written as a standard paper.
Include an appendix that lists and groups 20 to 30 relevant tools / papers
you found into the 3 to 5 "clusters" you identified.
Provide an URL or clear reference for the relevant tools / papers you found.
Hand-in
Post slides and report online and submit URL via dropbox.
Example
Financial Data Visualizations by Brian Sylvester: Powerpoint | Paper

How to Guide
What to Do
Specify Data Domain to be visualized.
Describe Steps and InfoVis Tool(s) to be used to accomlish specific task(s).
Create Screencasts and/or Tutorial to enable others to visualize a specific data domain.
Create Presentation
Create set of slides that showcase the screencasts and/or tutorials you have created.
Provide demonstration of working example.
Create Report
20 to 25 pages, written as a standard paper.
Can consist of tutorial contents and sceenshots of screencasts.
Hand-in
Post slides and report online and submit URL via dropbox.
Example
Visualizing Data: A Tutorial by Angelo Rousos: Part 1 | Part 2
How to Develop Visualizations from Social Media Analytics by Jamie Cann.

Usability Evaluation
What to Do
Select information visualization tool to evaluate.
Describe and motivate evaluation design.
Conduct evaluation with 3 to 5 people.
Each evaluation session should last 30 to 45 minutes.
Explain if the subjects received any training and, if yes, what it consisted of.
Video Record each session - you can use a digital video camera or 30 day free trial of Camtasia that let's you record the screen as well as the built-in video camera.
Have each subject sign the Informed Consent Form and collect Feedback & Suggestions after each session is completed.
Create Presentation
Create a set of slides that explain your usability design and your major findings.
Digitize TWO video clips of moments during the Usability Tests you want to share with the class.
Insert the two video clips on two separate slides.
Create Report
20 to 25 pages, written as a standard paper.
Describe major findings and suggestions.
Include an appendix with the signed Informed Consent forms.
Hand-in
Post slides and report online and submit URL via dropbox.
Example
TouchGraph Evaluation by Kelli Staley: Powerpoint | Report

Prototype
Resources
Powerpoint Tips and Presence Prototype example.
What to Do
Motivate domain choice.
Perform task and need analysis.
Describe design approach and information visualization principles used.
Develop prototype.
Have an "domain expert" use the prototype and provide feedback.
Create Presentation
Create a set of slides that describe task analysis and your design approach.
Include URL of prototype.
Report on the "domain expert" feedback.
Create Report
20 to 25 pages, written as a standard paper.
Provide screenshots of prototype and explain design approach.
Hand-in
Post slides and report online and submit URL via dropbox.
Example
Visualization Prototype and Initial Skteches and Paper by Cynthia Lovett

(Group) Data Story Project
What to Do
Select data domain to be visualized and identify data sources to be aggregated.
Develop group strategy for how best to visualize data and develop DataStory.
Assign tasks to team members.
Create visualizations and DataStory.
Describe group approach and insights gained.
Create Presentation
Create a set of slides that capture the key elements of the developed DataStory, approach taken and insights gained.
Create Report
Create DataStory as a set of web pages with embedded visualizations and describe approach taken and insights gained.
Hand-in
Post DataStory online and submit URL via dropbox.
Example
Visualizing 50 Years of Hit Songs by Shaun Ellis & Tom Engelhardt: Site
Media Coverage: Tableau Blog | DailySwarm | ChartPorn | NBC