Construction and Performance Optimization of a Multimodal Transformer-Based Fake News Detection Model
Innovative Applications of AI · 31.03.2026 · 3 (1) · 24-29
DOI: 10.70695/iaai202601a7 ↗Kurz gefasst
Fachartikel in der Zeitschrift Innovative Applications of AI, online erschienen am 31.03.2026. Verfasst von Qifan Huang.
Abstract (Originalsprache)
To address the issues of current false news characterized by multimodal fusion dissemination and the limited interpretability of existing methods, we explore a multimodal transformer detection model (SEL-MSIT) that integrates supervised contrastive learning and multi-stage interaction. Based on feature extraction, the supervised contrastive learning module enhances feature discrimination capability, while the multi-scale cross-modal interaction module is constructed to explore deep correlations. A consistent attention mechanism is employed to achieve efficient feature fusion. Experimental results demonstrate that SEL-MSIT outperforms mainstream baselines in terms of accuracy, precision, recall, and F1-score. Ablation experiments are conducted to verify the effectiveness of each optimization module, and the results can serve as supplementary data for decision-making.
Zitieren
APA
Huang, Q. (2026). Construction and Performance Optimization of a Multimodal Transformer-Based Fake News Detection Model. Innovative Applications of AI, 3(1), 24-29. https://doi.org/10.70695/iaai202601a7
Vancouver
Huang Q. Construction and Performance Optimization of a Multimodal Transformer-Based Fake News Detection Model. Innovative Applications of AI. 2026;3(1):24-29. doi:10.70695/iaai202601a7
BibTeX
@article{Huang2026,
author = {Huang, Qifan},
title = {Construction and Performance Optimization of a Multimodal Transformer-Based Fake News Detection Model},
journal = {Innovative Applications of AI},
year = {2026},
volume = {3},
number = {1},
pages = {24--29},
doi = {10.70695/iaai202601a7}
}
EUVIMED