University of Wisconsin-Madison scientists have developed a way for New Yorker magazine to use crowd-sourcing to judge captions in its weekly cartoon caption contest.
Engineering professor Rob Nowak and his team developed algorithms that let internet users instantaneously sort through some 5,000 submissions each week that a New Yorker staffer used to spend eight hours trawling through, CNET reports.
"The caption contest is intended for engagement," cartoon editor Bob Mankoff told CNET. "There's no reason why engagement shouldn't continue with which ones you think are best."
The adaptive crowdsourcing algorithms the New Yorker is using attempt to quickly weed out the weakest caption so more people will vote on potential winners. The magazine went public with the tool last month.
Nowak thinks it can be used in many other selective functions.
He wants to develop an adaptive algorithm to direct scientists to the lab tests most likely to net the needed results, saving money by shaving hours off research projects.
"It's kind of like the contest," Nowak said. "You really want to find the winners or the really important things, rather than just the good things."
And if the research could help teach computers what’s funny, the cartoon contest will wind up having significant impact on how people interface with their electronic devices.