One Word, Two Sides: Traces of Stance in Contextualized Word Representations

Aina Garí Soler; Matthieu Labeau; Chloé Clavel

One Word, Two Sides: Traces of Stance in Contextualized Word Representations

Aina Garí Soler
, Matthieu Labeau
, Chloé Clavel

Institut Polytechnique de Paris

Research output: Contribution to journal › Conference article › peer-review

Abstract

The way we use words is influenced by our opinion. We investigate whether this is reflected in contextualized word embeddings. For example, is the representation of “animal” different between people who would abolish zoos and those who would not? We explore this question from a Lexical Semantic Change standpoint. Our experiments with BERT embeddings derived from datasets with stance annotations reveal small but significant differences in word representations between opposing stances.

Original language	English
Pages (from-to)	3950-3959
Number of pages	10
Journal	Proceedings - International Conference on Computational Linguistics, COLING
Volume	29
Issue number	1
Publication status	Published - 1 Jan 2022
Event	29th International Conference on Computational Linguistics, COLING 2022 - Hybrid, Gyeongju, Korea, Republic of Duration: 12 Oct 2022 → 17 Oct 2022

Cite this

@article{8b584c9c10ba4102b12c61927e0db190,

title = "One Word, Two Sides: Traces of Stance in Contextualized Word Representations",

abstract = "The way we use words is influenced by our opinion. We investigate whether this is reflected in contextualized word embeddings. For example, is the representation of “animal” different between people who would abolish zoos and those who would not? We explore this question from a Lexical Semantic Change standpoint. Our experiments with BERT embeddings derived from datasets with stance annotations reveal small but significant differences in word representations between opposing stances.",

author = "Soler, \{Aina Gar{\'i}\} and Matthieu Labeau and Chlo{\'e} Clavel",

note = "Publisher Copyright: {\textcopyright} 2022 Proceedings - International Conference on Computational Linguistics, COLING. All rights reserved.; 29th International Conference on Computational Linguistics, COLING 2022 ; Conference date: 12-10-2022 Through 17-10-2022",

year = "2022",

month = jan,

day = "1",

language = "English",

volume = "29",

pages = "3950--3959",

journal = "Proceedings - International Conference on Computational Linguistics, COLING",

issn = "2951-2093",

number = "1",

}

TY - JOUR

T1 - One Word, Two Sides

T2 - 29th International Conference on Computational Linguistics, COLING 2022

AU - Soler, Aina Garí

AU - Labeau, Matthieu

AU - Clavel, Chloé

PY - 2022/1/1

Y1 - 2022/1/1

N2 - The way we use words is influenced by our opinion. We investigate whether this is reflected in contextualized word embeddings. For example, is the representation of “animal” different between people who would abolish zoos and those who would not? We explore this question from a Lexical Semantic Change standpoint. Our experiments with BERT embeddings derived from datasets with stance annotations reveal small but significant differences in word representations between opposing stances.

AB - The way we use words is influenced by our opinion. We investigate whether this is reflected in contextualized word embeddings. For example, is the representation of “animal” different between people who would abolish zoos and those who would not? We explore this question from a Lexical Semantic Change standpoint. Our experiments with BERT embeddings derived from datasets with stance annotations reveal small but significant differences in word representations between opposing stances.

UR - https://www.scopus.com/pages/publications/85153472244

M3 - Conference article

AN - SCOPUS:85153472244

SN - 2951-2093

VL - 29

SP - 3950

EP - 3959

JO - Proceedings - International Conference on Computational Linguistics, COLING

JF - Proceedings - International Conference on Computational Linguistics, COLING

IS - 1

Y2 - 12 October 2022 through 17 October 2022

ER -

One Word, Two Sides: Traces of Stance in Contextualized Word Representations

Abstract

Other files and links

Fingerprint

Cite this