Raising public awareness through the analysis of pre-election political discourse in Greece, by AUTh

Raising public awareness through the analysis of pre-election political discourse in Greece, by AUTh

IANUS partner, Aristotle University of Thessaloniki and its Data and Web Science Laboratory (DataLab) collaborated with the incubator for Media Education and Development Laboratory (iMedD Lab) in order to create a platform that analyses political speeches in Greece during the pre-election period (March – June 2023). One of its most important goals has been to raise public awareness on politics issues in Greece, by providing new insights on the underlying meanings of political discourse.

This project has been an experimental collaboration between professionals from journalism, political theory, data science, and AI. This collaboration aims to analyse campaign speeches delivered by the leaders of Greek political parties and examine the limits of modern technology, while also exploring the societal impact of such a project. Many technologies have been utilised in this work, including the AI chatbot ChatGPT.

In more detail, the platform depicts the results of sentiment analysis on the speeches of political leaders; it categorises sentiment into neutral, positive, negative, and tracks how speech sentiment can change throughout an entire pre-election period. Similarly, it presents the results of topic detection comparatively indicating which topics have been most frequently mentioned throughout leaders’ speeches. Finally, the platform detects whether the analysed political discourse features any instances of polarization and populism and provides statistics on the use of the word “will” (bearing a promising connotation) during the speeches.

Overall, this collaboration highlights the potential of technology and data-driven approaches in shedding light on political issues and complex societal phenomena, and in raising public awareness. Such analyses can help the public form their opinions and attitudes in an evidence-based and rigorous way.

The platform is publicly available in the following link: https://lab.imedd.org/en/elections-2023/

SideMenu
Scroll to Top