Divergent Emotional Patterns in Disinformation on Social Media? An Analysis of Tweets and TikToks about the DANA in Valencia
Iván Arcos, Paolo Rosso, Ramón Salaverría
TL;DR
This study tackles disinformation diffusion during the DANA event in Valencia by analyzing 650 posts from X and TikTok, combining emotion analysis, LIWC-based linguistic cues, and audio-feature analysis to characterize multimodal patterns. It compares traditional detectors (SVM+TF-IDF), transformer-based models with optional audio input, and GPT-4o Few-Shot labeling, finding platform-specific emotional signals ( sadness and fear on X; anger and disgust on TikTok) and demonstrating that audio-augmented models can surpass text-only baselines. The SVM+TF-IDF approach achieves the highest F1 when data are scarce, while incorporating audio features into roberta-large-bne and leveraging GPT-4o Few-Shot yields strong accuracy, underscoring the value of multimodal detection for TikTok. Overall, the work highlights how emotional and linguistic cues, augmented by audio signals, improve disinformation detection and offers a practical blueprint for real-time, multimodal misinformation monitoring across social media.
Abstract
This study investigates the dissemination of disinformation on social media platforms during the DANA event (DANA is a Spanish acronym for Depresion Aislada en Niveles Altos, translating to high-altitude isolated depression) that resulted in extremely heavy rainfall and devastating floods in Valencia, Spain, on October 29, 2024. We created a novel dataset of 650 TikTok and X posts, which was manually annotated to differentiate between disinformation and trustworthy content. Additionally, a Few-Shot annotation approach with GPT-4o achieved substantial agreement (Cohen's kappa of 0.684) with manual labels. Emotion analysis revealed that disinformation on X is mainly associated with increased sadness and fear, while on TikTok, it correlates with higher levels of anger and disgust. Linguistic analysis using the LIWC dictionary showed that trustworthy content utilizes more articulate and factual language, whereas disinformation employs negations, perceptual words, and personal anecdotes to appear credible. Audio analysis of TikTok posts highlighted distinct patterns: trustworthy audios featured brighter tones and robotic or monotone narration, promoting clarity and credibility, while disinformation audios leveraged tonal variation, emotional depth, and manipulative musical elements to amplify engagement. In detection models, SVM+TF-IDF achieved the highest F1-Score, excelling with limited data. Incorporating audio features into roberta-large-bne improved both Accuracy and F1-Score, surpassing its text-only counterpart and SVM in Accuracy. GPT-4o Few-Shot also performed well, showcasing the potential of large language models for automated disinformation detection. These findings demonstrate the importance of leveraging both textual and audio features for improved disinformation detection on multimodal platforms like TikTok.