Paper: Automatic Committed Belief Tagging

ACL ID C10-2117
Title Automatic Committed Belief Tagging
Venue International Conference on Computational Linguistics
Session Poster Session
Year 2010
Authors

We go beyond simple propositional mean- ing extraction and present experiments in determining which propositions in text the author believes. We show that deep syn- tactic parsing helps for this task. Our best feature combination achieves an F- measure of 64%, a relative reduction in F- measure error of 21% over not using syn- tactic features.