Papers for

speech analysis tool developers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

BERT model predicts Austrian politicians’ parties from speeches

Analysing Austrian Parliamentary Speeches by Predicting Party Affiliations

Abstract: This paper examines speech characteristics of the Austrian National Council members (CMs). We introduce analysis methodology based on a fine-tuned BERT-Model which assigns to every speech a probability distribution over Austrian parties. These are then aggregated at the speaker and party level. We perform an extensive evaluation to ensure the quality of the generated distributions prior to further analysis. Instead of considering mis-assignments of predicted parties as a performance issue, we leverage them to derive insights about the CMs and political developments. Our findings show that parties that govern together become more similar in their speech characteristics, while, analogously, parties that are in opposition move further away from governing parties. Moreover, CMs who switch party usually adapt their speech characteristics only after the switch occurs. Finally, we publicly release a data set containing party affiliation predictions for the speeches of the Austrian National Council from 1996 to 2018.

Fri 23 OctComputation and LanguageArtificial IntelligenceComputers and Society

The gist

This paper looks at how politicians in Austria speak and tries to figure out which political party they belong to using a computer model called BERT. Instead of just trying to get the party right, the authors use the model’s answers to learn about how political parties’ speaking styles change over time. They found that parties that work together start to sound more alike, and politicians sound different if they switch parties but only after they actually switch. The authors also share a big set of their predictions for speeches from 1996 to 2018 to help others study this.

Open → 2610.32515v1