Showing 1–1 of 1 results for author: da Silva, B C C

Enriching GNNs with Text Contextual Representations for Detecting Disinformation Campaigns on Social Media

Authors: Bruno Croso Cunha da Silva, Thomas Palmeira Ferraz, Roseli De Deus Lopes

Abstract: Disinformation on social media poses both societal and technical challenges, requiring robust detection systems. While previous studies have integrated textual information into propagation networks, they have yet to fully leverage the advancements in Transformer-based language models for high-quality contextual text representations. This work addresses this gap by incorporating Transformer-based t… ▽ More Disinformation on social media poses both societal and technical challenges, requiring robust detection systems. While previous studies have integrated textual information into propagation networks, they have yet to fully leverage the advancements in Transformer-based language models for high-quality contextual text representations. This work addresses this gap by incorporating Transformer-based textual features into Graph Neural Networks (GNNs) for fake news detection. We demonstrate that contextual text representations enhance GNN performance, achieving 33.8% relative improvement in Macro F1 over models without textual features and 9.3% over static text representations. We further investigate the impact of different feature sources and the effects of noisy data augmentation. We expect our methodology to open avenues for further research, and we made code publicly available. △ Less

Submitted 22 November, 2024; v1 submitted 24 October, 2024; originally announced October 2024.

Comments: Work still in progress. Accepted as Extended Abstract Poster at LoG Conference 2024