Register for an account

X

Enter your name and email address below.

Your email address is used to log in and will not be shared or sold. Read our privacy policy.

X

Website access code

Enter your access code into the form field below.

If you are a Zinio, Nook, Kindle, Apple, or Google Play subscriber, you can enter your website access code to gain subscriber access. Your website access code is located in the upper right corner of the Table of Contents page of your digital edition.

Technology

Fun with Google Correlate

Gene ExpressionBy Razib KhanMay 30, 2011 9:45 AM

Newsletter

Sign up for our email newsletter for the latest science news

A few people have pointed me to Google Correlate. Google says that this tool is like "Google Trends in reverse." You know I love Google Trends, so of course I'm going to poke around Google Correlate. The tool shows you the strongest correlations by query over time, as well as concentrations and correlations of the query by state. So here's the distribution for "Jesus":

correl1.png

If you know something about the United States, I assume you aren't surprised. But how about something like "Krishna?"

correl2.png

Does this make sense to you? What's up with New Jersey? As a proportion it has by far the greatest number of Indian Americans. (the #1 correlate is "Radha," which makes sense) How about something more obscure? OK, "Ludwig von Mises":

correl3.png

The Ludwig von Mises Institute is based out of Auburn, Alabama. They're a radical libertarian outfit whose aim is to promote the thought of the Austrian economist Ludwig von Mises (whose more moderate student, F. A. Hayek, is much more famous today). It doesn't seem like they're doing that great of a job at broader outreach about von Mises' legacy if Alabama, presumably the institute and its affiliate scholars, can make such a large splash on the query stream. But onto more pressing concerns in relation to the topicality of this weblog. For the "theory of evolution" most of the correlations make sense, but "the sumerians" and "hector of troy"? Either there's a lot of noise below a really high threshold of state level correlations, or, the world is more mysterious than we can comprehend. "Genetics" and "Genomics" were pretty straightforward. So far I've been looking at correlations of queries as they vary across states. Spatial. How about time series? "Theory of evolution" now is a little less weird. But why does it correlate so strongly with "jim crow laws"? Looking at the time series gives us a plausible hypothesis:

correl4.png

The school year. People are looking for "theory of evolution" and "jim crow laws" when school is in session. During the summer, and over Christmas vacation, they stop looking. Compare to "bank" and "bank of":

correl5.png

Finally, let's look at two different queries going in different directions. First, "restaurant":

correl6.png

And now, "groceries":

correl7.png

Finally, let's look at the trend for Facebook. First, the time series isn't surprising:

correl8.png

correl9.png

But look at the map which shows the concentration of queries. Facebook started at Harvard and other elite schools in the mid-2000s. But in 2011 the only place to expand its mindshare is in places like Maine and Mississippi I guess.

    3 Free Articles Left

    Want it all? Get unlimited access when you subscribe.

    Subscribe

    Already a subscriber? Register or Log In

    Want unlimited access?

    Subscribe today and save 70%

    Subscribe

    Already a subscriber? Register or Log In