Come hither or go away? Recognising pre-electoral coalition signals in the news

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 7798–7810 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 7798 Come Hither or Go Away? Recognising Pre-Electoral Coalition Signals in the News Ines Rehbein University of Mannheim Simone Paolo Ponzetto University of Mannheim {ines,simone,heiner}@informatik.uni-mannheim.de Heiner Stuckenschmidt University of Mannheim Anna Adendorf University of Mannheim {anna.adendorf,oke.bahnsen}@mzes.uni-mannheim.de Oke Bahnsen University of Mannheim Lukas Stoetzer Humboldt University Berlin lukas.stoetzer@hu-berlin.de Abstract In this paper, we introduce the task of politi- cal coalition signal prediction from text, that is, the task of recognizing from the news coverage leading up to an election the (un)willingness of political parties to form a government coali- tion. We decompose our problem into two re- lated, but distinct tasks: (i) predicting whether a reported statement from a politician or a jour- nalist refers to a potential coalition and (ii) predicting the polarity of the signal – namely, whether the speaker is in favour of or against the coalition. For this, we explore the benefits of multi-task learning and investigate which setup and task formulation is best suited for each sub-task. We evaluate our approach, based on hand-coded newspaper articles, cov- ering elections in three countries (Ireland, Ger- many, Austria) and two languages (English, German). Our results show that the multi-task learning approach can further improve results over a strong monolingual transfer learning baseline. 1 Introduction In the last decade the capabilities of computa- tional methods for text analysis have drastically evolved and the usage of Natural Language Process- ing (NLP) methods have increasingly expanded in scope, enabling a wide range of applications for the automatised analysis of political texts (Glavaš et al., 2019). Much work in NLP in the past has fo- cused on making sense of how legislative processes are codified in text – e.g., by analysing debate mo- tions (Abercrombie et al., 2019; Abercrombie and Batista-Navarro, 2020; Sawhney et al., 2020, inter alia). However, when analysing electoral dynamics it seems necessary to complement the application of NLP methods on domain-specific texts such as electoral manifestos with a broader analysis of sig- nals from news texts (Blokker et al., 2020). Specifically in the case of government forma- tion in multi-party systems – that is, in countries where the government is typically formed by mul- tiple parties – party members and candidates often communicate in interviews and rallies during elec- tion campaigns which coalition their party might or might not be willing to enter after the election. These public statements about prospects of future governments are often picked up by media outlets. Being able to automatically identify coalition sig- nals in news coverage has accordingly the potential to enable large-scale studies of voting behaviour using ‘text-as-data’, with the promise of improv- ing our understanding of voters’ behaviour and of strategic voting in multi-party democratic sys- tems (Bahnsen et al., 2020; Gschwend et al., 2017; Stoetzer and Orlowski, 2020). Very recent work in political science made first steps in this direc- tion by investigating dictionary-based approaches to automatically …

Content truncated to 500 words