Social Perceptions of Language Variation: Comparing Humans and LLMs: DiLCo Lecture Series 2024 (23 May) : Digital language variation in context (DiLCo)

Link:
  • https://lecture2go.uni-hamburg.de/l2go/-/get/v/69252
Autor/in:
Beteiligte Person:
  • Regionales Rechenzentrum der Universität Hamburg/ MCC/ Lecture2Go
Verlag/Körperschaft:
Universität Hamburg
Erscheinungsjahr:
2024
Medientyp:
Audiovisuell
Schlagworte:
  • language variation
  • LLM
  • computational linguistics
  • sociolinguistics
  • social media
  • digital linguistics
  • ChatGPT
  • Sprache, Literatur, Medien (SLM I + II)
Beschreibung:
  • Dong Nguyen is an assistant professor at the Department of Information and Computing Sciences at Utrecht University. At Utrecht University, she is heading the Natural Language Processing (NLP) and Society Lab. She is interested in the intersection between NLP and sociolinguistics, exploring how NLP can be used to support sociolinguistic research and how sociolinguistics can inform the development of better NLP models. There are often various ways to express the same thing. Think of, for example, the different words we can use for a given concept, or the many creative spellings in social media. Language variation is often seen as a challenge for developing robust NLP models. However, language variation is also an opportunity: it can help us develop models that are more sensitive to social context. In this talk, I will present current research into the associations that Large Language Models form regarding spelling variations and how these associations align with human perceptions. DiLCo Lecture Series 2024 aims to showcase cutting edge international research on digitally language and communication by both senior and younger researchers from across the world. We wish to present research that explores digital language and communication by drawing on key concepts and topics in socio-cultural linguistics, such as community, context, identity, mediated interaction, multimodality, and linguistic change. We particularly welcome presentations of innovative methods that cut across traditional disciplinary boundaries.
Beziehungen:
URL https://lecture2go.uni-hamburg.de/l2go/-/get/l/6829
Lizenz:
  • UHH-L2G
Quellsystem:
Lecture2Go UHH
Interne Metadaten
Quelldatensatz
oai:lecture2go.uni-hamburg.de:69252