Visualizing Words and Emotions

Great story by Assma Malik in the Montreal Gazette on the data visualization work of Jonathan Harris.

“In 2004, computer scientist and artist Jonathan Harris created a data visualization called WordCount. The interactive presentation shows the 86,800 most popular English words, ranked and sized in order of how often they’re used, and arranged side by side in one long, scrollable sentence.

Speaking at the Online News Association conference, Harris said he found the ranking of words by frequency of use led to some pretty profound juxtapositions. The word “god” (No. 376) is one word from “began,” two words from “start” and six words from “war.” And one string of words (ranked Nos. 992-995) reads “America, ensure, oil, opportunity.”

But while these seemingly random phrases hidden in the words were striking, Harris found what users did with his WordCount tool to be even more revealing.

He had built a search function that allowed them to look for a particular word and its ranking. And when he pulled the most popular queries, he discovered that the word most often searched was “sex,” followed by “the” and “love.” A little further down came some choice four-letter words, and then the word “god” (now No. 7). He called this new site, which strung together the most frequently searched words on WordCount, QueryCount.

With its data drawn from the British National Corpus, a 100-million-word collection of written and spoken language meant to represent English usage, WordCount itself was codified and rigid. But QueryCount showed that people’s interaction with it made the information more fluid. They were able to make their mark on the data through their selections.

“If you design a system that is simple enough,” Harris explained, “people will find a way to use that for human expression.”

Harris took his fascination with human expression further, co-creating the We Feel Fine project, a data visualization that searches the blogosphere for phrases that describe emotion ( “I feel like …”) and categorizes them according to age, location, feeling, gender, weather and date.

The result is a beautiful interactive experience with star-like dots of every colour and size representing human emotion. If you click on a blue one, it reads “I feel like I am watching everything spin around me and I can’t ground myself.” If you click on a yellow one, “I feel good and my husband says I look good.” You don’t see the words within the context of the blog entry, though you see the age, gender and the location of the writer. Their feelings stand alone in a constellation of universal emotion…..

……His approach to telling human stories through techniques often used for cold, hard data is similar to using information to tell effective news stories. He says the way he sorts his emotional data focuses on universal storytelling elements: “characters, concepts, contexts, colours and excitement level.” He has just released a book featuring moments and photos from We Feel Fine.

Harris sorts through endless data streams, mining for those nuggets of collective truth.

“The way you deal with chaos,” he said, “is you define a good lens.” Sounds like something every news organization would do well to focus on.

About GilPress

I launched the Big Data conversation; writing, research, marketing services; &
This entry was posted in data visualization. Bookmark the permalink.

1 Response to Visualizing Words and Emotions

  1. Jean Gogolin says:

    I loved this one, Gil! Fascinating stuff.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s