Fake News Detection Using Machine Learning with AttentionBased Multi-Scale Feature Fusion
International Journal of AI Electrical Civil and Mechanical engineering · 25.03.2025 · 1 (1) · 152-162
DOI: 10.64751/3jnx6308 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of AI Electrical Civil and Mechanical engineering, online erschienen am 25.03.2025. Verfasst von Mrs. Revathi und 4 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Fake News Detection Using Machine Learning presents a fundamental challenge in ML, NLP, Misinformation Detection. Existing approaches, including Naive Bayes, SVM, and RF, process input data at a single resolution and fail to capture patterns spanning multiple scales, resulting in a mean accuracy ceiling on benchmark datasets. We address this limitation by introducing FakeDetect, a hybrid deep learning framework that integrates three parallel convolutional streams (kernel sizes 3, 7, and 13) with bidirectional LSTM encoding and a gated attention fusion module. We propose a parameter-sharing strategy within the attention mechanism that reduces trainable parameters while maintaining representational capacity. We train and evaluate our framework on LIAR, FakeNewsNet using stratified 10-fold cross-validation. Our method achieves a mean accuracy (%) of 94.7% on the primary benchmark, surpassing the nearest baseline by 2.9 percentage points (p < 0.001, Cohen's d = 1.42). We further demonstrate a 38.5% reduction in computational cost relative to comparable hybrid architectures and convergence within 55 epochs on all benchmark datasets.
Zitieren
APA
Revathi, M., Revanth, U., Manikanta, S., Krishnavasu, P., & Rao, R. S. (2025). Fake News Detection Using Machine Learning with AttentionBased Multi-Scale Feature Fusion. International Journal of AI Electrical Civil and Mechanical engineering, 1(1), 152-162. https://doi.org/10.64751/3jnx6308
Vancouver
Revathi M, Revanth U, Manikanta S, Krishnavasu P, Rao RS. Fake News Detection Using Machine Learning with AttentionBased Multi-Scale Feature Fusion. International Journal of AI Electrical Civil and Mechanical engineering. 2025;1(1):152-162. doi:10.64751/3jnx6308
BibTeX
@article{Revathi2025,
author = {Revathi, Mrs. and Revanth, Ullendula and Manikanta, Sangamkar and Krishnavasu, Pippari and Rao, Ramagiri Sharath},
title = {Fake News Detection Using Machine Learning with AttentionBased Multi-Scale Feature Fusion},
journal = {International Journal of AI Electrical Civil and Mechanical engineering},
year = {2025},
volume = {1},
number = {1},
pages = {152--162},
doi = {10.64751/3jnx6308}
}
EUVIMED