EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

1511. Predictive Politics: Understanding Drivers of Engagement in Belgian Politics

Invited abstract in session WB-31: Learning Analytics and other Text Analytics tasks, stream Analytics.

Wednesday, 10:30-12:00
Room: 046 (building: 208)

Authors (first author is the speaker)

1. Dylan Van Mulders
Marketing, Innovation and Organisation, Ghent University
2. Matthias Bogaert
Marketing, Innovation and Organization, Ghent University
3. Dirk Van den Poel
MIO/Data Analytics, Ghent University

Abstract

Social media platforms have offered politicians and political parties new opportunities to communicate and interact with citizens. Especially in Belgium, political parties and politicians are spending a lot of money to social network sites to attract voters. Backed by Belgian party financing, Belgian parties have been among the biggest spenders in Europe, even outside of campaign periods.
The main goal of this research is to gain a deeper understanding on the online engagement with political posts within the Belgian multi-party system. We focus on what politicians are talking about and how people engage with this on X, formerly known as Twitter, since it remains one of the most political social network sites to date. To achieve this, we propose a two-step approach to predict the level of engagement based on a unique set of features. This approach first classifies tweets into several like-minded categories, and then applies regression on each category. Supported by the Academic Twitter API we analyzed the posts of 72 Belgian politicians as well as posts from all the major political party Twitter profiles for a period of 2 years.
The main contribution of this study is that it is the first to combine the added value of multiple distinct characteristics in the context of online engagement and disclose differences between posts originating from personal and party profiles within a multi-party political system.

Keywords

Status: accepted

Back to the list of papers