Skip to main content

Vocabulary of natural language processing

Search from vocabulary

Concept information

Preferred term

linguistic category  

Broader concept

Example

  • Each of the layers is specified by its relevant linguistic category (stem paradigm case number gender etc.) and possibly by an error label. (Rosen, 2016)
  • The fact that some linguistic categories have a steeper curve than the others may also signalise the difficulty of these categories from a machine learning perspective. (Stadler, Macketanz & Avramidis, 2021)
  • To study whether different linguistic categories are jointly encoded within LLMs we use a crossneutralization method. (Starace, Papakostas, Choenni, Panagiotopoulos, Rosati, Leidinger & Shutova, 2023)
  • We study information sharing between linguistic categories in LLMs finding evidence of joint encoding between pairs of related POS tag classes. (Starace, Papakostas, Choenni, Panagiotopoulos, Rosati, Leidinger & Shutova, 2023)
  • We test this hypothesis by cross-neutralizing every linguistic category in a language A with another category from a language B e.g. neutralizing all Italian POS tags with English nouns. (Starace, Papakostas, Choenni, Panagiotopoulos, Rosati, Leidinger & Shutova, 2023)

In other languages

URI

http://data.loterre.fr/ark:/67375/8LP-WFW61X72-6

Download this concept:

RDF/XML TURTLE JSON-LD Last modified 5/31/24