What is VerbNet?
VerbNet (VN) (Kipper-Schuler 2006) is the largest on-line verb lexicon currently available for English. It is a hierarchical domain-independent, broad-coverage verb lexicon with mappings to other lexical resources such as WordNet (Miller, 1990; Fellbaum, 1998), Xtag (XTAG Research Group, 2001), and FrameNet (Baker et al., 1998). VerbNet is organized into verb classes extending Levin (1993) classes through refinement and addition of subclasses to achieve syntactic and semantic coherence among members of a class. Each verb class in VN is completely described by thematic roles, selectional restrictions on the arguments, and frames consisting of a syntactic description and semantic predicates with a temporal function, in a manner similar to the event decomposition of Moens and Steedman (1988).
(return to top)
The VerbNet Approach
VerbNet groups together verbs with identical sets of syntactic frames and semantic predicate structures, as in the example below for class CUT-21.1-1:
These classes may inherit frames from a parent class, resulting in a hierarchical structure:
The classes in VerbNet are based on an extension of Levin (1993) classes, with the understanding that the syntactic form of a verb and its arguments informs its semantics. Observe, for example, the differences and similarities between the following two classes, CUT-21.1 and BREAK-45.1:
(return to top)
How It All Works
Each VN class contains a set of syntactic descriptions, or syntactic frames, depicting the possible surface realizations of the argument structure for constructions such as transitive, intransitive, prepositional phrases, resultatives, and a large set of diathesis alternations. Semantic restrictions (such as animate, human, organization) are used to constrain the types of thematic roles allowed by the arguments, and further restrictions may be imposed to indicate the syntactic nature of the constituent likely to be associated with the thematic role. Syntactic frames may also be constrained in terms of which prepositions are allowed. Each frame is associated with explicit semantic information, expressed as a conjunction of boolean semantic predicates such as `motion,' `contact,' or `cause.' Each semantic predicate is associated with an event variable E that allows predicates to specify when in the event the predicate is true (start(E) for preparatory stage, during(E) for the culmination stage, and end(E) for the consequent stage). Figure 1. shows a complete entry for a frame in VerbNet class Hit-18.1.
VerbNet has recently been integrated with 57 new classes from Korhonen and Briscoe's (2004) (K&B) proposed extension to Levin's original classification (Kipper et al., 2006). This work has involved associating detailed syntactic-semantic descriptions to the K&B classes, as well as organizing them appropriately into the existing VN taxonomy. An additional set of 53 new classes from Korhonen and Ryant (2005) (K&R) have also been incorporated into VN. The outcome is a freely available resource which constitutes the most comprehensive and versatile Levin-style verb classification for English. After the two extensions VN has now also increased our coverage of PropBank tokens (Palmer et. al., 2005) from 78.45% to 90.86%, making feasible the creation of a substantial training corpus annotated with VN thematic role labels and class membership assignments, to be released in 2007. This will finally enable large-scale experimentation on the utility of syntax-based classes for improving the performance of syntactic parsers and semantic role labelers on new domains.
Integrating the two recent extensions to Levin classes into VerbNet was an important step in order to address a major limitation of Levin's verb classification, namely the fact that verbs taking ADJP, ADVP, predicative, control and sentential complements were not included or addressed in depth in that work. This limitation excludes many verbs that are highly frequent in language. A summary of how this integration affected VN and the result of the extended VN is shown in Table 1. The figures show that our work enriched and expanded VN considerably. The number of first-level classes grew significantly (from 191 to 274), there was also a significant increase in the number of verb senses and lemmas, along with the set of semantic predicates and the syntactic restrictions on sentential complements.
Each verb argument is assigned one (usually unique) thematic role within the class. A few exceptions to this uniqueness are classes which contain verbs with symmetrical arguments, such as Chitchat-37.6 class, or the ContiguousLocation-47.8 class. These classes have indexed roles such as Actor1 and Actor2, as explained above.
(return to top)