EUVIMEDEUVIMED entdecken
Crossref

Do all politicians sound the same? Comparing model explanations to human responses

Otto Tarkka, Kimmo Elo, Filip Ginter, Veronika Laippala

Digital Humanities Quarterly · 20.02.2026 · 20 (1)

DOI: 10.63744/vjurh6rtug2p ↗

Kurz gefasst

Fachartikel in der Zeitschrift Digital Humanities Quarterly, online erschienen am 20.02.2026. Verfasst von Otto Tarkka und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

It is sometimes said that all politicians sound the same with their speeches mired in political jargon full of clichés and false promises. To investigate how distinct the plenary speeches of political parties truly are and what linguistic features make them distinct, we trained a BERT classifier to predict the party affiliation of Finnish members of parliament from their plenary speeches. We contrasted and compared model performance to human responses to see how humans and the model differ in their ability to distinguish between the parties. We used the model explainability method SHAP to identify the linguistic cues that the model most relies on. We show that a deep learning model can distinguish between parties much more accurately than the respondents to the questionnaire. The SHAP explanations and questionnaire responses reveal that whereas humans tend to rely on mostly topical cues, the model has learned to recognize other cues as well, such as personal style and rhetoric.

Zitieren

APA

Tarkka, O., Elo, K., Ginter, F., & Laippala, V. (2026). Do all politicians sound the same? Comparing model explanations to human responses. Digital Humanities Quarterly, 20(1). https://doi.org/10.63744/vjurh6rtug2p

Vancouver

Tarkka O, Elo K, Ginter F, Laippala V. Do all politicians sound the same? Comparing model explanations to human responses. Digital Humanities Quarterly. 2026;20(1). doi:10.63744/vjurh6rtug2p

BibTeX
@article{Tarkka2026,
  author = {Tarkka, Otto and Elo, Kimmo and Ginter, Filip and Laippala, Veronika},
  title = {Do all politicians sound the same? Comparing model explanations to human responses},
  journal = {Digital Humanities Quarterly},
  year = {2026},
  volume = {20},
  number = {1},
  doi = {10.63744/vjurh6rtug2p}
}
Kostenlos starten →