VerbNet (Kipper-Schuler 2005) is a hierarchical domain- independent, broad-coverage verb lexicon with mappings to several widely-used verb resources, including WordNet (Miller 1990, Fellbaum 1998), Xtag (Xtag 2001), and FrameNet (Baker et al. 1998). It includes syntactic and semantic information for classes of English verbs derived and then extended from Levin's original classification (Levin 1993). The original Levin classes have been refined and new subclasses added to achieve syntactic and semantic coherence among members.
VerbNet is available at Unified Verb Index Reference Page.
Funded by GALE
The PropBank annotates a large corpus of text labeled with semantic propositions. The predicate-argument relations or semantic role labels are added as a semantic layer over the syntactic trees produced by the Penn Treebank. PropBank is available through LDC. The PropBank Frame Files have entries for individual verbs that provide guidance to the annotators. They currently cover 99% of verbs in general English news text(newswire, talk shows, webtext, etc.). As such they provide an invaluable lexical resource. They are also available at the Unified Verb Index Reference Page. Contact us for downloadable versions.
Ontonotes Sense Groups
Funded by GALE
The OntoNotes Sense Groups project provides groupings of fine-grained WordNet senses into more coarse-grained senses. WordNet (Miller 1990, Fellbaum 1998) uses fine-grained distinctions that provide subtle information about the particular usages of various lexical items. When used as a resource for annotation of various genres of text, this fine level of granularity has not been conducive to high automatic tagging performance. Annotation of verb senses as described by coarse-grained PropBank framesets may result in higher performance, but the blurring of distinctions between verb senses with similar argument structures may fail to alleviate the problems posed by ambiguity. Our goal in this project is to provide verb sense resource at a middle level of granularity that allow us to capture as much information as possible from a lexical item while still attaining high system performance in automatic sense disambiguation. OntoNotes Sense Groupings are also viewable at the Unified Verb Index Reference Page.