Paper: Situated Models of Meaning for Sports Video Retrieval

ACL ID N07-2010
Title Situated Models of Meaning for Sports Video Retrieval
Venue Human Language Technologies
Session Short Paper
Year 2007
Authors

Situated models of meaning ground words in the non-linguistic context, or situation, to which they refer. Applying such models to sports video re- trieval requires learning appropriate representa- tions for complex events. We propose a method that uses data mining to discover temporal pat- terns in video, and pair these patterns with associ- ated closed captioning text. This paired corpus is used to train a situated model of meaning that sig- nificantly improves video retrieval performance.