• search hit 1 of 1
Back to Result List

Time series influences in political communication

  • Current political issues are often reflected in social media discussions, gathering politicians and voters on common platforms. As these can affect the public perception of politics, the inner dynamics and backgrounds of such debates are of great scientific interest. This thesis takes user generated messages from an up-to-date dataset of considerable relevance as Time Series, and applies a topic-based analysis of inspiration and agenda setting to it. The Institute for Web Science and Technologies of the University Koblenz-Landau has collected Twitter data generated beforehand by candidates of the European Parliament Election 2019. This work processes and analyzes the dataset for various properties, while focusing on the influence of politicians and media on online debates. An algorithm to cluster tweets into topical threads is introduced. Subsequently, Sequential Association Rules are mined, yielding wide array of potential influence relations between both actors and topics. The elaborated methodology can be configured with different parameters and is extensible in functionality and scope of application.

Download full text files

Export metadata

Author:Tobias Thesing
Referee:Steffen Staab, Oul Han, Sarah de Nigris
Document Type:Master's Thesis
Date of completion:2019/12/07
Date of publication:2019/12/11
Publishing institution:Universität Koblenz, Universitätsbibliothek
Granting institution:Universität Koblenz, Fachbereich 4
Date of final exam:2019/12/07
Release Date:2019/12/11
Tag:2019 European Parliament Election; Association Rules; Political Communication
Number of pages:78
Institutes:Fachbereich 4 / Institute for Web Science and Technologies
Dewey Decimal Classification:0 Informatik, Informationswissenschaft, allgemeine Werke / 00 Informatik, Wissen, Systeme / 004 Datenverarbeitung; Informatik
3 Sozialwissenschaften / 32 Politikwissenschaft / 320 Politikwissenschaft
BKL-Classification:54 Informatik / 54.08 Informatik in Beziehung zu Mensch und Gesellschaft
Licence (German):License LogoEs gilt das deutsche Urheberrecht: § 53 UrhG