Social Media Data Mining as a Novel Approach to Electoral Geography
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
When individual discourses and attitudes on social media are used for electoral geography with their spatio-temporal information, they can provide essential clues about voter behavior and voting preferences. Although there are different election-based studies using data from social media, these studies have remained far from spatial understanding of the reasons behind the votes. This research aims to understand the differences in voter behavior and voting, along with revealing the context effects on voting by using this new data and methods in the field of electoral geography. We obtain the election-oriented tweets on X (Twitter) posted from Yenimahalle (Turkey), determine sentiment states, identify the dominant topics in the tweets, and reveal the relationship between the popularity rates of the candidates on X (Twitter) and the actual vote rates. Consequently, user-generated spatio-temporal data on X (Twitter) is essential in understanding the differences in voter behavior/participation and voting, as well as in revealing context effects on voting.