Paper: Automatic Classification of English Verbs Using Rich Syntactic Features

ACL ID I08-2107
Title Automatic Classification of English Verbs Using Rich Syntactic Features
Venue International Joint Conference on Natural Language Processing
Session Main Conference
Year 2008
Authors
  • Lin Sun (University of Technology of Troyes, Troyes France; University “A.I. Cuza”, Iasi Romania)
  • Anna Korhonen (Cambridge University, Cambridge UK)
  • Yuval Krymolowski (University of Haifa, Haifa Israel)

Previous research has shown that syntactic features are the most informative features in automatic verb classification. We exper- iment with a new, rich feature set, extracted from a large automatically acquired subcate- gorisation lexicon for English, which incor- porates information about arguments as well as adjuncts. We evaluate this feature set us- ing a set of supervised classifiers, most of which are new to the task. The best classi- fier (based on Maximum Entropy) yields the promising accuracy of 60.1% in classifying 204 verbs to 17 Levin (1993) classes. We discuss the impact of this result on the state- of-art, and propose avenues for future work.