Daniel Hardt

4 paper cards in Mova's library · 1997–2021

Daniel Hardt wrote 4 of the papers Mova reads, in 1997–2021, most of them on Coreference resolution, Discourse analysis; often with Anders Søgaard, Matthew Lamm, Owen Rambow.

The profile is the library's match of the name; two people with one name are not always told apart.

Papers

Ellipsis Resolution as Question Answering: An Evaluation (2021)
The paper treats ellipsis resolution as a reading-comprehension problem. The elided clause becomes a question, and the antecedent becomes the answer span in the preceding text. The authors train BERT-based models on sluice ellipsis and VP ellipsis, first alone and then jointly with question-answering and coreference…
Antecedent Selection for Sluicing: Structure and Content (2016)
The paper asks which earlier sentence or clause a sluice refers to. It argues that both syntax and discourse matter, and it builds features for the structural relation between the sluice and each candidate, for discourse relations, and for the match between the candidate's content and the wh-type of the sluice. A…
Generation of VP Ellipsis: A Corpus-Based Approach (2001)
The paper asks when a generator should elide a verb phrase and leave an auxiliary in its place, as in "in 1987, only 9% did".
An Empirical Approach to VP Ellipsis (1997)
Hardt builds the first corpus-based system for resolving VP ellipsis (VPE). The data are 644 VPE examples found in the parsed Penn Treebank, 260 of them from the Wall Street Journal. A syntactic filter first removes candidate antecedents in impossible configurations. The remaining candidates are ranked by four…