The game is named "Page Hunt" and the basic idea is to show the player a random web page, and get the player to come up with a query that would bring up this page in the top few results on a search engine. The idea is that human computation games that elicit data from players can be used to improve search.
Page Hunt was developed by Raman Chandrasekar and Chris Quirk of Microsoft Research in partnership with The Chinese University of Hong Kong and Georgia Institute of Technology. According to Microsoft "before we released the game to the web, we conducted a pilot study within our company. During this experiment, 341 people played Page Hunt over a period of ten days, generating over 14,400 labels on 744 web pages in the system. On average every player contributed 42 labels, and every page has about 19 labels."
Page Hunt was unveiled at this week's SIGIR 2009 conference in Boston. If you're interested in helping improve Bing search results then checkout Page Hunt which is hosted at Microsoft's Live Labs site.
No comments:
Post a Comment