Studying Speed Dates to Predict Social Intention

We all subconsciously analyze the meaning of words beyond their literal translations of words, but what if there were factors we could use to predict things like social intention? That’s exactly what Dan Jurafsky’s research aimed to do. He and his colleagues studied almost 1,000 speed dates to gain insight into how to extract social meaning. They found that by using computers and certain predictors they could detect social intentions with above 70 percent accuracy — far better than any human. The only problem we have watching this lecture is the amount of times Jurafsky uses the word “um.” It gets a little frustrating after awhile. But if you can get past it, the research is really very fascinating.

Keep up with new videos daily through RSS or Email