Indexat a

Llicència i ús

Citacions

Altmetrics

Grant support

This work has been funded by Grant PLEC2021-007681 (XAIDisInfodemics) and PID2020-117263GB-100 (FightDIS) funded by MCIN/AEI/10.13039/501100011033 and by "ERDF A way of making Europe", by the "European Union" or by the "European Union NextGenerationEU/PRTR", by grant PCI2022-134990-2 (MARTINI) of the CHISTERA IV Cofund 2021 program, funded by MCIN/AEI/10.13039/501100011033 and by the "European Union NextGenerationEU/PRTR", by Calouste Gulbenkian Foundation, under the project MuseAI - Detecting and matching suspicious claims with AI.

Anàlisi d'autories institucional

Compartir

Multimodal Visio-Lingual Content Analysis to Detect Fake Content on Reddit

Publicat a:Lecture Notes In Computer Science. 15346 143-154 - 2025-01-01 15346(), DOI: 10.1007/978-3-031-77731-8_14

Autors: Giron, Adrian; Huertas-Tato, Javier; Camacho, David

Afiliacions

Resum

The spread of misinformation across online social networks poses a significant threat to political stability, societal harmony, and economic integrity. This paper tackles the challenge of analyzing online posts, which often consist of diverse heterogeneous modalities like text or image. We introduce a multimodal architecture designed to accurately identify manipulated content. Our approach leverages a large-scale, multimodal dataset encompassing images, captions, comments, and metadata for each post. A novel encoding strategy is employed to capture both the semantic content and hierarchical structure of the comments. Based on CLIP, an early fusion technique is applied to process and merge the hidden representations. This method effectively combines multimodal and unimodal data according to their respective information channels. The performance is evaluated against the Fakeddit dataset, achieving a competent accuracy in binary classification tasks (0.9506 acc(test)) with significant hardware limitations, and outperforming the SotA in more complex categorization tasks, with 0.9509 acc(test) and 0.9371 acc(test) in 3-way and 6-way label classification, respectively.

Paraules clau