Bold predictions are made every day. We’ll reduce our carbon emissions by 50% in 20 years, boast business leaders. No, make that 80% in 15 years. We’ll cut the deficit in half by 2015, pandering politicians claim. That leaves us with dozens of conflicting estimates and ballpark figures that are soon forgotten. It’s hard to hold experts to their predictions, but that could all change soon thanks to an experimental search engine from Yahoo.
Developed by the company’s Barcelona research lab, Time Explorer is a search engine for the past, present, and future. Results are displayed on a timeline that stretches years back and forward. Move your mouse over the future part of the timeline, and you get predictions for what was supposed to happen in that year from as much as 20 years ago. For example, the timeline for “North Korea” lets us know that the rogue state should have developed some 200 nuclear warheads–according to an inaccurate op-ed in the New York Times by Nicholas Kristof in 2004.
So far, the search engine has been seeded with 1.8 million New York Times articles stretching back to 1987. That means it isn’t just useful for keeping up on prediction accuracy, it’s also a fascinating demonstration of how topics evolve over time. Results for “Bill Clinton” in the late ’80s return local reports from Arkansas; by the mid-’90s, opinions on his first presidency; in the aughts, features on his various public initiatives. All in all, the experiment offers a time-stamped portrait of people, places, and events, which is lost in the overwhelming hodgepodge of results returned on your typical search engine.
“You can see, for wars or any other event, not only the people that are important, but when they became important,” Michael Matthews, a member of the Yahoo research team, told MIT’s Technology Review. “The evolution of news over time is not something you can do very easily with tools that are out there today.” (Take that, Google!)
Yahoo’s experiment is an interesting departure from the typical direction of search engine betas. Today, search result “relevancy” often hinges on how recently the site has been updated. Google and Bing are working tirelessly to perfect their “real-time” engines, which enable users to see what’s happening now by archiving every blog post or tweet, regardless of significance. The Barcelona lab’s attempt to gauge relevance across a wide spectrum of time, including the future, is a refreshing approach to search technology.
And just wait until they start archiving Nostradamus’ works.