Word embeddings on ideology and issues from Swedish parliamentarians’ motions : a comparative approach
Fredén, Annika LU
; Johansson, Moa
and Saynova, Denitsa
(2024)
In Journal of Elections, Public Opinion and Parties
- Abstract
Quantitative analysis of large-scale political text data in the form of word embeddings has great potential for systematising differences between political parties. We examine the differences between embeddings obtained from speakers from the two competitors for the PM position in Sweden (Social Democrats and Moderates) over a 30-year period. The goal is to compare how off-the-shelf general pre-trained models perform relative to pre-training on a smaller dataset from the same domain. In the analysis, we focus on two types of concepts: issues and ideological terms. We find that generally, the off-the-shelf pre-trained models lead to more reliable results and greater emphasis on ideological differences between the studied parties.
- Abstract (Swedish)
- Quantitative analysis of large-scale political text data in the form of word embeddings has great potential for systematising differences between political parties. We examine the differences between embeddings obtained from speakers from the two competitors for the PM position in Sweden (Social Democrats and Moderates) over a 30-year period. The goal is to compare how off-the-shelf general pre-trained models perform relative to pre-training on a smaller dataset from the same domain. In the analysis, we focus on two types of concepts: issues and ideological terms. We find that generally, the off-the-shelf pre-trained models lead to more reliable results and greater emphasis on ideological differences between the studied parties.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/72519bc3-b82a-4a32-8ed1-a147e9aeeeb6
- author
- Fredén, Annika
LU
; Johansson, Moa
and Saynova, Denitsa
- organization
- publishing date
- 2024-12-04
- type
- Contribution to journal
- publication status
- in press
- subject
- keywords
- machine learning, Parliaments, text as data, word embeddings, word embeddings, parties, ideology
- in
- Journal of Elections, Public Opinion and Parties
- publisher
- Routledge
- external identifiers
-
- scopus:85210919347
- ISSN
- 1745-7289
- DOI
- 10.1080/17457289.2024.2433979
- project
- Bias and methods of AI technology studying political behavior
- language
- English
- LU publication?
- yes
- additional info
- Publisher Copyright: © 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
- id
- 72519bc3-b82a-4a32-8ed1-a147e9aeeeb6
- date added to LUP
- 2025-01-28 14:00:54
- date last changed
- 2025-05-06 20:38:57
@article{72519bc3-b82a-4a32-8ed1-a147e9aeeeb6,
abstract = {{<p>Quantitative analysis of large-scale political text data in the form of word embeddings has great potential for systematising differences between political parties. We examine the differences between embeddings obtained from speakers from the two competitors for the PM position in Sweden (Social Democrats and Moderates) over a 30-year period. The goal is to compare how off-the-shelf general pre-trained models perform relative to pre-training on a smaller dataset from the same domain. In the analysis, we focus on two types of concepts: issues and ideological terms. We find that generally, the off-the-shelf pre-trained models lead to more reliable results and greater emphasis on ideological differences between the studied parties.</p>}},
author = {{Fredén, Annika and Johansson, Moa and Saynova, Denitsa}},
issn = {{1745-7289}},
keywords = {{machine learning; Parliaments; text as data; word embeddings; word embeddings; parties; ideology}},
language = {{eng}},
month = {{12}},
publisher = {{Routledge}},
series = {{Journal of Elections, Public Opinion and Parties}},
title = {{Word embeddings on ideology and issues from Swedish parliamentarians’ motions : a comparative approach}},
url = {{http://dx.doi.org/10.1080/17457289.2024.2433979}},
doi = {{10.1080/17457289.2024.2433979}},
year = {{2024}},
}