Abstract
Verb phrase (VP) ellipsis is the omission of a verb phrase whose meaning can be reconstructed from the linguistic or real-world context. It is licensed in English by auxiliary verbs, often modal auxiliaries: She can go to Hawaii but he can’t [e]. This paper describes a system called ViPER (VP Ellipsis Resolver) that detects and resolves VP ellipsis, relying on linguistic principles such as syntactic parallelism, modality correlations, and the delineation of core vs. peripheral sentence constituents. The key insight guiding the work is that not all cases of ellipsis are equally difficult: some can be detected and resolved with high confidence even before we are able to build systems with human-level semantic and pragmatic understanding of text.
Highlights
Verb phrase (VP) ellipsis is the omission of a VP whose meaning can be reconstructed from the linguistic or real-world context
For reasons that will become clear shortly, it divides hits into those with a hard discourse break, a soft discourse break, and a conjunction. This method will not offer full recall since VP ellipsis need not be followed by a punctuation mark or conjunction; it covers many cases and was more than sufficient to support the testing of our linguistic hypotheses about ellipsis resolution
Its output is (a) a set of examples that it believes include VP ellipsis prior to a hard discourse break, a soft discourse break, or a conjunction; (b) the subset of examples that it believes it can treat with reasonable confidence; (c) ellipsis resolutions for the latter; and (d) metadata indicating which resolution strategy was applied to each example, for purposes of testing and debugging both the implementation and the knowledge supporting it
Summary
Verb phrase (VP) ellipsis is the omission of a VP whose meaning can be reconstructed from the linguistic or real-world context. ViPER pursues string-level resolution of VP ellipsis, defined as copying the sponsor string to fill the elliptical gap This is a midstream result toward full meaning analysis; and, as a midstream result, it has. Example (4) is syntactically simple because the clause containing the sponsor occurs right before the ellipsis clause and there are no additional main verbs competing to be the sponsor. ViPER carries out modality analysis to determine how the modal verbs in sponsor clauses should be handled during ellipsis resolution. The remainder of the paper describes how ViPER detects ellipsis (Section 2), selects which instances it can treat (Section 3), and determines the actual sponsor once the sponsor clause has been identified (Section 4). We report a system evaluation (Section 5), place this contribution within the larger field (Section 6), and describe ongoing and future work (Section 7)
Published Version (Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have