Northwestern University and University of Minnesota computer scientists have just created a website that you might spend the rest of your day exploring.
It's called Atlasify, and it displays search results through "heat maps" that signifies information about a given topic in a geographic area. For example, if you searched "Beer" a world map would pop up and every country would be colored in a different shade of green, darker meaning there is more connection between beer and that country (if you're wondering, Europe was almost entirely dark green, though the US, Australia, and New Zealand were also a deep shade of green). Red dots signify a special spot of information about that topic.
You can then click on a country and find out more information about how that location and the topic interact. For example, clicking on Bulgaria (one of the darkest green shade countries in Europe) you would find that Kamenitza is one of the top beer brands, and had 18 percent of the Bulgarian beer market in 2005 (you're welcome trivia nerds).
The site is in beta testing right now, so users can give feedback about the relevance of information. The researchers said the site can work for over 4.8 million queries. You can also choose a random search topic (possibilities: Harry Potter, Microbrewery, Lenovo).
But that's not all. You can also see your results in the form of a periodic table, senate seats, or timeline. The periodic table form is the most developed right now and can yield some fascinating results. Search "World War II" for example, and the darkest element (literally, figuratively) is uranium, given it's connection to nuclear bombs.
The senate seat and timeline model are still alpha testing, so there is less activity but definitely has potential. A search for "technology" shows which senators have sat on the commerce, science, and transportation committee.
The site was developed in collaboration between social computing researchers at GroupLens at the University of Minnesota and WebSAIL at Northwestern.
One of the researchers at GroupLens, Brent Hecht, explained a bit more about how the algorithms work in a blog post:
"To generate these maps and explanations, Atlasify uses a family of algorithms that we call “semantic relatedness plus explanation” measures (SR+E measures, for short) and a technique we call “explicit spatialization”, an homage to the explicit semantic analysis technique from Evgeniy Gabrilovich (now at Google) and colleagues. If you’re interested in learning more, we’ve published the full details in a paper at SIGIR."
This is a fascinating innovation on its own, but also can be a valuable tool for understanding how something functions in through a more complex and dynamic lens than just a list of word search results. For example, if you'll scroll down you'll see the "geography of virtual reality." I noticed that Sweden was a faint shade of green, so I clicked to find out more, and discovered a Swedish artist who once had museum goers attend a virtual exhibit in which they could run around and destroy art.
A new perspective on VR indeed.
Naturally we wanted to search all things technology and innovation. Here's a look at what we found:
Geography of technology:
Chemistry of technology:
Geography of innovation:
Geography of mobile phones:
Geography of virtual reality
Image credits: Screenshots via Atlasify
