Referentiality in individual named event embeddings

Gosse Minnema, Aurélie Herbelot

Research output: Contribution to conferencePosterAcademic

8 Downloads (Pure)

Abstract

Distributional models of meaning are known to be good at capturing conceptual information about generic concepts, but it is unclear to what extent they can also capture referential information about individual entities. Events are particularly difficult to model distributionally because of their large diversity in linguistic forms (they could be expressed as verbs, nominalizations, common nouns, or even be completely implicit), and because it is unclear what should serve as the basis of a distributional representation of an individual event: bare verbs, predicate-argument structures, or even whole sentences? Here, inspired by previous work proposing distributional models for entity-denoting proper names (e.g., “Angela Merkel”, “Barcelona”) (Gupta et al., 2015; Herbelot, 2015), we propose using event-denoting proper names (“Hurricane Sandy”, “Battle of Waterloo”, “The Paul McCartney World Tour”) as a starting point for investigating individual events.
Original languageEnglish
Publication statusPublished - May-2020
Externally publishedYes
EventGeCKo Symposium: Integrating Generic and Contextual Knowledge - (online), Barcelona, Spain
Duration: 18-May-202018-May-2020
https://sites.google.com/view/gecko2020

Workshop

WorkshopGeCKo Symposium
CountrySpain
CityBarcelona
Period18/05/202018/05/2020
Internet address

Cite this