Dan Jurafsky : 2010 Plenary Session

Automatically detecting social intentions from spoken conversation is an important task for social computing, and key for building conversational agents.We describe a system that detects whether a speaker is awkward, friendly, or flirtatious with above 70% accuracy, significantly outperforming not only the baseline but also, for flirtation, outperforming the human interlocutors. We find that features like pitch and the use of emotional vocabulary help detect flirtation, collaborative conversational style (laughter, questions) help in detecting friendliness, and disfluencies help in detecting awkwardness. In analyzing why our system outperforms humans, we show that humans are very poor perceivers of flirtatiousness, and instead often project their own intended behavior onto their interlocutors.

This talk describes joint work with Dan McFarland (School of Education) and Rajesh Ranganath (Computer Science Department).

Dan Jurafsky is a Professor of Linguistics, and by courtesy of Computer Science, at Stanford University. From 1996-2003 he was on the faculty of the University of Colorado, Boulder. Dan received a B.A in Linguistics in 1983 and a Ph.D. in Computer Science in 1992, both from the University of California at Berkeley, and was a postdoc 1992-1995 at the International Computer Science Institute in Berkeley. He is the recipient of a 2002 MacArthur Fellowship, and is the co-author with Jim Martin of the widely-used textbook "Speech and Language Processing". Recent research interests include the automatic extraction of conceptual and social meaning, machine translation, the similarities and differences between human and machine language processing, the application of natural language processing to the social sciences, and the linguistics of food.