6th ANNUAL COMPTEXT Conference 2024

Abstract:

This study explores the effects of manipulated visual content on polarized climate change debates on Twitter by answering the following research question: Do ‘real’ vs manipulated images about climate change on Twitter lead to different levels of engagement and interactions between believers and skeptics (deniers)? In this research, we analyse the intersection of social science and computer science in the context of climate change discourse on Twitter by adopting a multimodal and computational approach combining automated image and text analysis to examine more than 700,000 images, and replies shared by Twitter users in the year 2019. By using hash functions, and a “lab-scenarios” approach commonly used in computer science to detect manipulated images, we retrieve pairs of near-identical images, and analyse the manipulation between the real and the manipulated image within each pair. Then, we use different computational techniques such as topic modelling (BERTopic) and Latent Semantic Scaling (LSS) to process and explore the Tweet texts and comments. Results reveal user differences in the distribution of engagement between manipulated and real images. These differences concern not only the type of visual content engaged by deniers and believers but also how these users react to it regarding valence and topic prevalence. Here, we generally see that the believers exhibit more engagement than deniers when they are exposed to real images. This research contributes to understanding the role of imagery in the climate change debate, introducing a novel and multimodal approach that combines advanced visual detection with multiple text analysis techniques.

My Presentation: