Diversity and language technology: how language modeling bias causes epistemic injustice

It is well known that AI-based language technology—large language models, machine translation systems, multilingual dictionaries, and corpora—is currently limited to three percent of the world’s most widely spoken, financially and politically backed languages. In response, recent efforts have sought...

Full description

Saved in:

Bibliographic Details
Published in:	Ethics and information technology Vol. 26; no. 1; p. 8
Main Authors:	Helm, Paula, Bella, Gábor, Koch, Gertraud, Giunchiglia, Fausto
Format:	Journal Article
Language:	English
Published:	Dordrecht Springer Netherlands 01.03.2024 Springer Nature B.V Springer Verlag
Subjects:	Artificial Intelligence Bias Bilingual dictionaries Community Computation and Language Computer Science Computers and Society Dialects Dictionaries Ethics History, Philosophy and Sociology of Sciences Humanities and Social Sciences Innovation/Technology Management Justice Language Language modeling Languages Large language models Library Science Machine translation Management of Computing and Information Systems Marginality Modelling Original Paper Sociolects Speech communities Technology Underserved populations User Interfaces and Human Computer Interaction Language technology Large language models Epistemic injustice Diversity Language modeling bias Linguistic bias Digital divide Lexical gaps epistemic injustice language modeling bias bias linguistic diversity
ISSN:	1388-1957, 1572-8439
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Be the first to leave a comment!