IBKA-MSM: A Novel Multimodal Fake News Detection Model Based on Improved Swarm Intelligence Optimization Algorithm,...

Area of Science:

• Artificial Intelligence
• Computer Science
• Data Science

Background:

• Social media misinformation presents complex challenges due to diverse modalities and cross-semantic correlations.
• Accurate detection of fake news is hindered by semantic inconsistency and uneven modality dependency.

Purpose of the Study:

• To propose a novel multimodal semantic representation framework, IBKA-MSM, for enhanced fake news detection.
• To address the critical challenge of detecting misinformation under conditions of semantic inconsistency and uneven modality dependency.

Main Methods:

• Developed the IBKA-MSM framework integrating swarm intelligence optimization with deep neural modeling.
• Employed an Improved Black-Winged Kite Algorithm (IBKA) for feature selection, featuring adaptive step-size control and enhanced memory mechanisms.
• Introduced Modality-Generated Loop Verification (MGLV) for semantic alignment and a Semantic Confidence Matrix with Modality-Coupled Interaction (SCM-MCI) for adaptive multimodal fusion.

Main Results:

• IBKA-MSM achieved a high accuracy of 95.80%, outperforming existing mainstream hybrid models.
• The F1 score showed significant improvement: approximately 2.8% over Particle Swarm Optimization (PSO) and 1.6% over basic Black-Winged Kite Algorithm (BKA).

Conclusions:

• The IBKA-MSM framework demonstrates robustness and strong capability in maintaining multimodal semantic consistency for fake news detection.
• The proposed approach effectively addresses the complexities of multimodal misinformation, offering a significant advancement in detection accuracy and reliability.