Predicting elections for multiple countries using Twitter and polls

Tsakalidis, Adam; Papadopoulos, S.; Cristea, A.I.; Kompatsiaris, Yiannis

Abstract

The authors' work focuses on predicting the 2014 European Union elections in three different countries using Twitter and polls. Past works in this domain relying strictly on Twitter data have been proven ineffective. Others, using polls as their ground truth, have raised questions regarding the contribution of Twitter data for this task. Here, the authors treat this task as a multivariate time-series forecast, extracting Twitter- and poll-based features and training different predictive algorithms. They've achieved better results than several past works and the commercial baseline.

Citation

Tsakalidis, A., Papadopoulos, S., Cristea, A., & Kompatsiaris, Y. (2015). Predicting elections for multiple countries using Twitter and polls. IEEE Intelligent Systems, 30(2), 10-17. https://doi.org/10.1109/mis.2015.17

• Journal Article Type: Article
• Acceptance Date: Jan 5, 2015
• Online Publication Date: Jan 26, 2015
• Publication Date: Mar 1, 2015
• Deposit Date: Jul 11, 2018
• Publicly Available Date: Jul 31, 2018
• Journal: IEEE Intelligent Systems
• Print ISSN: 1541-1672
• Electronic ISSN: 1941-1294
• Publisher: Institute of Electrical and Electronics Engineers
• Peer Reviewed: Peer Reviewed
• Volume: 30
• Issue: 2
• Pages: 10-17
• DOI: https://doi.org/10.1109/mis.2015.17
• Public URL: https://durham-repository.worktribe.com/output/1326470
• Related Public URLs: http://wrap.warwick.ac.uk/75812/