Triple Path Enhanced Neural Architecture Search for Multimodal Fake News Detection

2025-01-24

AI Summary

To address two key challenges in multimodal fake news detection on social media—rigid modality fusion and poor generalization under partial modality missing—this paper proposes the first tri-path searchable architecture: two dynamic paths for adaptive processing of unimodal or bimodal inputs, and one static path to model deep cross-modal semantic correlations. Leveraging neural architecture search (NAS), we design an end-to-end differentiable optimization framework integrating modality-adaptive gating, heterogeneous path co-training, and multi-granularity feature alignment. Evaluated on multiple benchmark datasets, our method consistently outperforms state-of-the-art approaches, achieving absolute accuracy gains of 3.2–5.8 percentage points in image-text missing scenarios. These results demonstrate superior robustness and generalization capability, especially under real-world data incompleteness.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Multimodal fake news detection has become one of the most crucial issues on social media platforms. Although existing methods have achieved advanced performance, two main challenges persist: (1) Under-performed multimodal news information fusion due to model architecture solidification, and (2) weak generalization ability on partial-modality contained fake news. To meet these challenges, we propose a novel and flexible triple path enhanced neural architecture search model MUSE. MUSE includes two dynamic paths for detecting partial-modality contained fake news and a static path for exploiting potential multimodal correlations. Experimental results show that MUSE achieves stable performance improvement over the baselines.

Problem

• Fake News Detection
• Model Rigidity
• Partial Information Handling

Innovation

• MUSE model
• multi-path approach
• fake news detection