Researchers at Microsoft’s labs in Redmond, WA, have released an online game to help fine-tune search results.
Called Page Hunt, the game presents players with web pages and asks them to guess the queries that would produce the page within its first five results. Players score 100 points if the page is no.1 on the list, 90 points if it’s no.2, and so on. Bonuses are also awarded for avoiding frequently-used queries.
The idea is to gather useful information on user search habits which could be used to fine tune search algorithms and ranking scheme. The game was developed by Chris Quirk and Raman Chandrasekar at Microsoft, and colleagues from Georgia Tech and the Chinese University of Hong Kong, and it was unveiled this week at the SIGIR09 conference in Boston.
Page Hunt is a clever twist on “human computation”–using people to perform tasks that computers find difficult to do. Luis von Ahn, a professor at Carnegie Mellon University, has been a pioneer in this area, and has developed several similar projects: spam-fighting text puzzles that simultaneously help digitize old books, and games that help tag images and music with the relevant keywords. Another cool example of human computation in action is, of course, Amazon’s Mechanical Turk.
The researchers behind Page Hunt have already made one curious finding while testing the game internally: the longer a page’s URL (in characters), the harder it was for users to match the page to query words. The research don’t speculate about why this should be, but here’s a graph showing the relationship between URL length and the “findability” of a page:
I found Page Hunt
strangely addictive, although my first score was a pathetic 630.
A paper describing the Page Hunt research can be found here (pdf).
Smaller design teams can now prototype and deploy faster.