Abstract
Real-time large-scale data streams provided by Twitter create new possibilities for political scientists to nowcast political events. We developed a pipeline to process, analyze and aggregate data for presentation on a web application. During the 2021 German federal election campaign, expressed stances on competing political parties and their front-runners were analyzed in real time. State-of-the-art linguistic neural networks were reused and adapted by post-training and fine-tuning for detecting stances toward political actors. Furthermore, a dictionary-based approach was adopted to analyze the salient topics during the campaign across 32 (policy) issues. Within stance detection, a decrease in performance over time became visible, which can largely be attributed to a shift in issue focus on Twitter during the election campaign. This is emphasized with concrete empirical examples. During the final phase of the campaign, qualitative monitoring was maintained to ensure the validity and reliability of our findings. Based on this, potential error sources are presented, and possible solutions for future research are offered.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
The results presented here are part of the larger interdisciplinary research project SPARTA (Society, Politics and Risk with Twitter Analysis). The WebApp is accessible via: https://dtecbw.de/sparta?lang=en.
- 2.
Stance detection means inferring user’s attitude towards a predefined entity or topic. See Sect. 5 for details.
- 3.
In contrast to two-party systems, Germany has a more advanced multidimensional political space [12], although spatial models may work with a simplified left-right-scale.
- 4.
The 2021 election resulted in a coalition government made up of SPD, Bündnis 90/Die Grünen and FDP.
- 5.
The archived summary can be found at: https://dtecbw.de/sparta/germanelection.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
Original German tweet text: @c_lindner hat was an der Waffel. In einer Zeit wo Leute oft nur widerwillig Scholz oder Laschet wählen, und nach der Flut noch sauer auf Regierungsparteien, könnte FDP Kanzlerkandidat oder Kandidatin zw. 20–30% kommen. Wo ist der Mut der bei jedem Parteitag beschworen wird.
- 16.
Although there are many similar examples, according to our training dataset, the polarity assigned to these targets only differs in 1.7% of the cases.
- 17.
Disaster Control (Katastrophenschutz) is substantial during catastrophic events.
- 18.
NRW is the German federal state of North Rhine-Westphalia, where Armin Laschet was governor at the time.
- 19.
Triell was the TV debate between the leading candidates of CDU/CSU, SPD and Bündnis 90/Die Grünen.
- 20.
Original German tweet text: “Wir müssen mehr Sozialwohnungen bauen. Ich mache das in NRW.” sagt #Laschet im #Triell. Ist das so?
- 21.
Original German tweet text: Der Blockpartei und den gekauften Medien ist kein Mittel zu dreckig um die AfD in Misskredit zu bringen.
- 22.
Original German tweet text: Das Empörungsspektakel mit einem gefährlichen Gemisch aus Fakten und Unwahrheiten zeigt die Angst vor Machtverlust, da die Grünen einen Plan haben, wie sich Klimaschutz und Wohlstand verbinden lassen. #Kretschmann #btw21 #dBDK21.
References
Aldayel, A.: Stance detection on social media: state of the art and trends (2021). https://doi.org/10.1016/j.ipm.2021.102597
Barbieri, F., Anke, L.E., Camacho-Collados, J.: XLM-T: a multilingual language model toolkit for Twitter (2021). https://arxiv.org/abs/2104.12250v1
Bechini, A., Bondielli, A., Ducange, P., Marcelloni, F., Renda, A.: Addressing event-driven concept drift in twitter stream: a stance detection application. IEEE Access 9, 77758–77770 (2021)
Beck, P.A.: The electoral cycle and patterns of American politics. Br. J. Polit. Sci. 9(02), 129 (1979). https://doi.org/10.1017/S0007123400001691
Bundeswahlleiter: Heft 4 Wahlbeteiligung und Stimmabgabe nach Geschlecht und Altersgruppen. Wahl zum 20. Deutschen Bundestag am 26, 4 Sept 2021 (2022)
Ceron, A., Curini, L., Drews, W.: Short-term issue emphasis on twitter during the 2017 German election: a comparison of the economic left-right and socio-cultural dimensions. German Polit. 1–20 (2020). https://doi.org/10.1080/09644008.2020.1836161
Ceron, A., Curini, L., Iacus, S.M.: Politics and Big Data. Nowcasting and Forecasting Elections with Social Media. Routledge (2017)
Dai, J., Yan, H., Sun, T., Liu, P., Qiu, X.: Does syntax matter? A strong baseline for aspect-based sentiment analysis with RoBERTa (2021)
Darwish, K., Stefanov, P., Aupetit, M., Nakov, P.: Unsupervised user stance detection on Twitter. In: Proceedings of the International AAAI Conference on Web and Social Media, vol. 14, pp. 141–152 (2020)
Devlin, J., Chang, M.W., Lee, K., Toutanova, K.: BERT: pre-training of deep bidirectional transformers for language understanding (2019)
Dong, X., Lian, Y.: A review of social media-based public opinion analyses: challenges and recommendations. Technol. Soc. 67, 101724 (2021)
Downs, A.: An Economic Theory of Democracy. Harper and Row, New York (1957)
Duggan, M., Ellison, N., Lampe, C., Lenhart, A., Madden, M.: Demographics of key social networking platforms. Pew Research Center (2015). http://www.pewinternet.org/2015/01/09/demographics-of-key-social-networking-platforms-2/
Fraisier, O., Cabanac, G., Pitarch, Y., Besançon, R., Boughanem, M.: Stance classification through proximity-based community detection 18 (2018). https://doi.org/10.1145/3209542.3209549
Gao, Z., Feng, A., Song, X., Wu, X.: Target-dependent sentiment classification with BERT. IEEE Access 7, 154290–154299 (2019)
Ghosh, S., Singhania, P., Singh, S., Rudra, K., Ghosh, S.: Stance detection in web and social media: a comparative study (2020). https://github.com/prajwal1210/Stance-Detection-in-Web-and-Social-Media
Giorgioni, S., Politi, M., Salman, S., Croce, D., Basili, R.: UNITOR @ sardistance 2020: combining transformer-based architectures and transfer learning for robust stance detection (2020)
Göhring, A., Klenner, M., Conrad, S.: DeInStance: creating and evaluating a German corpus for fine-grained inferred stance detection (2021). https://huggingface.co/dbmdz/bert-base-german-cased
Guhr, O., Schumann, A.K., Bahrmann, F., Böhme, H.J.: Training a broad-coverage German sentiment classification model for dialog systems, pp. 11–16 (2020)
Gupta, Y., Kumar, P.: Real-time sentiment analysis of tweets: a case study of punjab elections. In: Proceedings of 2019 3rd IEEE International Conference on Electrical, Computer and Communication Technologies, ICECCT 2019 (2019). https://doi.org/10.1109/ICECCT.2019.8869203
Hamdi, A., et al.: A multilingual dataset for named entity recognition, entity linking and stance detection in historical newspapers (2021). https://doi.org/10.1145/3404835.3463255
Hardalov, M., Arora, A., Nakov, P., Augenstein, I.: Cross-domain label-adaptive stance detection (2021)
Hitesh, M.S., Vaibhav, V., Kalki, Y.J., Kamtam, S.H., Kumari, S.: Real-time sentiment analysis of 2019 election tweets using word2vec and random forest model. In: 2019 2nd International Conference on Intelligent Communication and Computational Techniques, ICCT 2019, pp. 146–151 (2019). https://doi.org/10.1109/ICCT46177.2019.8969049
Jungherr, A., Schoen, H., Jürgens, P.: The mediation of politics through Twitter: an analysis of messages posted during the campaign for the German federal election 2013. J. Comput. Mediated Commun. 21(1), 50–68 (2016). https://doi.org/10.1111/JCC4.12143
Kagan, V., Stevens, A., Subrahmanian, V.S.: Using Twitter sentiment to forecast the 2013 Pakistani election and the 2014 Indian election. IEEE Intell. Syst. 30(1), 2–5 (2015). https://doi.org/10.1109/MIS.2015.16
Klüver, H., Spoon, J.J.: Who responds? voters, parties and issue attention. Br. J. Polit. Sci. 46(3), 633–654 (2016)
Lehmbruch, G.: Parteienwettbewerb im Bundesstaat: Regelsysteme und Spannungslagen im Politischen System der Bundesrepublik Deutschland. VS Verlag für Sozialwissenschaften, Wiesbaden (2000)
Li, Y., Caragea, C.: A multi-task learning framework for multi-target stance detection, pp. 2320–2326 (2021)
Lijphart, A.: Patterns of Democracy. Government Forms and Performance in Thirty-Six Countries, Yale University Press, New Haven (2012)
Liu, Y., et al.: RoBERTa: a robustly optimized BERT pretraining approach (2019)
Mainwaring, S., Gervasoni, C., España-Najera, A.: Extra- and within-system electoral volatility. Party Polit. 23(6), 623–635 (2017)
Meier, F., Bazo, A., Elsweiler, D., Bazo, A.: Using social media data to analyse issue engagement during the 2017 German federal election (2021)
Pennebaker, J.W., Booth, R.J., Francis, M.E.: Linguistic inquiry and word count: Liwc [computer software], p. 135. liwc. net, Austin, TX (2007)
Rashed, A., Kutlu, M., Darwish, K., Elsayed, T., Bayrak, C.: Embeddings-based clustering for target specific stances: the case of a polarized Turkey (2020)
Reuver, M., Verberne, S., Morante, R., Fokkens, A.: Is stance detection topic-independent and cross-topic generalizable? - A reproduction study (2021). https://webis.de/events/sameside-19/
Roemmele, A., Gibson, R.: Scientific and subversive: the two faces of the fourth era of political campaigning. New Media Soc. 22(4), 595–610 (2020)
Samih, Y., Darwish, K.: A few topical tweets are enough for effective user stance detection, pp. 2637–2646 (2021)
Scharpf, F.W., Reissert, B., Schnabel, F.: Politikverflechtung: Theorie und Empirie des kooperativen Föderalismus in der Bundesrepublik. Monographien Ergebnisse der Sozialwissenschaften 1. Scriptor-Verl., Kronberg (1976)
Shi, T., Tech, V., Kang, K., Choo, J., Reddy, C.K.: Short-text topic modeling via non-negative matrix factorization enriched with local word-context correlations, p. 10 (2018). https://doi.org/10.1145/3178876.3186009
Soler, J.M., Cuartero, F., Roblizo, M.: Twitter as a tool for predicting elections results (2012). https://doi.org/10.1109/ASONAM.2012.206
Stier, S., Bleier, A., Lietz, H., Strohmaier, M.: Election campaigning on social media: politicians, audiences, and the mediation of political communication on Facebook and Twitter. Polit. Commun. 35(1), 50–74 (2018)
Sun, Q., Wang, Z., Li, S., Zhu, Q., Zhou, G.: Stance detection via sentiment information and neural network model. Front. Comput. Sci. 13(1), 127–138 (2019). https://doi.org/10.1007/s11704-018-7150-9
Tumasjan, A., Sprenger, T.O., Sandner, P.G., Welpe, I.M.: Predicting elections with twitter: what 140 characters reveal about political sentiment. In: Proceedings of the International AAAI Conference on Web and Social Media, vol. 4, no. 1, pp. 178–185 (2010)
Vamvas, J., Sennrich, R.: X-Stance: a multilingual multi-target dataset for stance detection (2020). https://doi.org/10.5281/zenodo.3831317
Wang, H., Can, D., Kazemzadeh, A., Bar, F., Narayanan, S.: A system for real-time Twitter sentiment analysis of 2012 U.S. presidential election cycle, pp. 8–14 (2012). http://t.co/qEns1Pmi
Wolf, T., et al.: Transformers: state-of-the-art natural language processing (2019). https://github.com/huggingface/
Xu, H., Liu, B., Shu, L., Yu, P.S.: BERT post-training for review reading comprehension and aspect-based sentiment analysis (2019)
Acknowledgements
This research is funded by dtec.bw—Digitalization and Technology Research Center of the Bundeswehr [project SPARTA]. We would like to thank all team members of SPARTA, in particular, the core team that worked so hard to make our real-time monitoring during the 2021 German election campaign possible: Andreas Neumeier, Benedikt Radtke, Martin Riedl and Johannes Steup.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 IFIP International Federation for Information Processing
About this paper
Cite this paper
Müller, A., Riedl, J., Drews, W. (2022). Real-Time Stance Detection and Issue Analysis of the 2021 German Federal Election Campaign on Twitter. In: Janssen, M., et al. Electronic Government. EGOV 2022. Lecture Notes in Computer Science, vol 13391. Springer, Cham. https://doi.org/10.1007/978-3-031-15086-9_9
Download citation
DOI: https://doi.org/10.1007/978-3-031-15086-9_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-15085-2
Online ISBN: 978-3-031-15086-9
eBook Packages: Computer ScienceComputer Science (R0)
