Video reference: A video question answering engine

Abstract

Community-based question answering systems have become very popular for providing answers to a wide variety of "how-to" questions. However, most such systems present only textual answers. In many cases, users would prefer visual answers such as videos which are more intuitive and informative. The Video Reference system is proposed as a solution to the above problem. It automatically extracts videos from YouTube as a video reference responding to a textual question. The demo shows results on real questions sampled from Yahoo! Answering. © 2010 Springer-Verlag Berlin Heidelberg.