YouTube For Your Data: Many Eyes on Obama & McCain

Is it possible to condense two weeks of soaring rhetoric by politicians into a single graphic? Let's find out.

In my last post I covered free online tools for creating information graphics.  Here is a look at another free online tool that can be used to easily create sophisticated visualizations and information graphics.


Is it possible to condense two weeks of soaring rhetoric by politicians into a single graphic?  The New York Times did just that in an efficient visualization of the speeches made during the Democratic and Republican conventions.  Titled "The Words They Used",  the graphic highlights and compares the themes that speakers both parties chose to focus on. The technique used in the NYT graphic is similar to visual technique popular on websites today known as a  "word cloud" or a "tag cloud". A word cloud is a visual representation of the frequency of words used in a body of text. (Some examples: Tagclouds at the social bookmarking site Delicious and on Planetizen Radar.)

Wouldn't it be useful if we could use this same technique to analyze other texts that we often encounter such as articles, reports, policies, plans, and websites? 

Let's start with Planetizen's news coverage. Here is a word cloud of Planetizen's news stories for the first half of this year:




I created this graphic using a free online tool called Wordle created by Jonathan Feinberg.  Creating word clouds couldn't be simpler using Wordle. Copy and paste some text and then click "go" to create your own word clouds.

Here is a word cloud of Planetizen news for the same time period filed under "environment" and "energy".

Now, let's use word clouds to look at John McCain's and Barack Obama's policy statements on energy, climate change, and the environment. Both candidates have published statements on their respective websites about their position on climate change, the environment, and energy.

To create word clouds of these statements, I used a similar web application developed by IBM's  research and development labs called "ManyEyes". More about Many Eyes later in this post. First, let's look at the word clouds.

(Click on the visualization graphic to see the interactive version.)

1. John McCain on Energy  
[ Source | Visualization ]



2. John McCain on Climate Change
[ Source | Visualization ]



3. Barack Obama on Energy & Environment
[ Source | Visualization ]




And now, a comparative word cloud.


4. McCain on Energy & Climate Change vs. Obama on Energy & Environment
[Sources same as above |  Visualization ]

The following visualization is called an "interleaved tag cloud". McCain's words are in brown and Obama's words are in blue.



[Note: The John McCain website has two separate statements on energy and climate change. The Barack Obama website has one statement on both environment and energy.  For the purposes of this demonstration, I combined the two McCain statements. Also, the source text on the two candidates websites may have changed since I created the visualizations.]

Barack Obama has also published an Urban Policy Plan. I could not find McCain's plan on the topic. Here are some visualizations of Obama's plan:

• Barack Obama Urban Policy Web Page Tagcloud
• Barack Obama Urban Policy Plan Tagcloud
• Barack Obama Urban Policy Plan: Transportation Word Tree
• Barack Obama Urban Policy Plan: Housing Word Tree

ManyEyes

ManyEyes was created by Martin Wattenberg and Fernanda B.Viégas of the Visual Communication Lab at IBM's Watson Research Center. Here is a video of Dr. Viegas explaining what you can do with Many Eyes.

Visualization tool: ManyEyes Interviewed by JD Lasica on Vimeo.

To summarize, you can upload your own data to Many Eyes. The data can be text as we did in the above examples. ManyEyes can handle the data you have sitting around in spreadsheets and databases too. You can upload numerical data in a specific format. You can then create visualizations from your data. Many Eyes offers 16 different types of visualizations at this time including maps, graphs, charts, histograms. treemaps, scatter plots, network diagrams, word clouds, and word trees.

Browse through the over 2100 visualizations already created at ManyEyes.  To learn how to create your own see Richard Heog's  tutorial.

ManyEyes aims to bring visualization tools within reach of more people by taking away the need for special computer software or skills. It is indeed easy to upload data and create visualizations that can help  analyze data and spotting trends or patterns. (It is also possible to create meaningless visualizations easily.)

This is where ManyEyes does something interesting.  By uploading data to ManyEyes users agree to share their data. So you can create visualizations from data uploaded by other users. Many Eyes also features "topic hubs" where users organize datasets and visualizations on a particular topic and get together to share insights.  This is similar to how websites such as Flickr and YouTube work -- by encouraging and make it easy for anyone to upload,  categorize, share, and comment on their digital content.  By letting "many eyes" look at and work with your data and visualizations,  there is the possibility of others spotting something that you haven't. 

We know this collaborative approach works for open source software, political blogs, and wikis. Will it work for data analysis and visualization?

[Also see Part 1 -- Infographics for the rest of us ]