Informace o publikaci

Building a Dataset for Detection of Verb Coordinations with a Shared Argument

Autoři

MEDKOVÁ Helena

Rok publikování 2021
Druh Článek ve sborníku
Konference Recent Advances in Slavonic Natural Language Processing (RASLAN 2021)
Fakulta / Pracoviště MU

Filozofická fakulta

Citace
www
Klíčová slova Coordination; Zeugma; Syntactic analysis; UDPipe 2; Brat annotation tool; VerbaLex
Popis Coordinate structures represent a specific linguistic problem relating to questions of sentence boundaries and multiple sentence element [1]. A particular di?iculty lies in processing at the level of automatic syntactic analysis of the sentence. To deal with the outlined issue, we decided to use the machine learning classification method, for which it is necessary to prepare a su?iciently large amount of data. This paper presents the methods and procedures we used to build a dataset focused on the phenomenon of verb coordinations that may share an argument in context.
Související projekty:

Používáte starou verzi internetového prohlížeče. Doporučujeme aktualizovat Váš prohlížeč na nejnovější verzi.

Další info