This article introduces a multimodal adversarial network to detect fake news in India, leveraging both text and image features. The method employs adversarial training to extract robust, discriminative characteristics for distinguishing authentic news from fake news. Evaluated on an Indian fake news dataset, the approach achieves high accuracy (0.89) and F1-score (0.90), demonstrating its effectiveness in identifying and mitigating misinformation in the Indian context. This method offers a significant step toward curbing fake news distribution on social media.

AUTHOR:  Dr. Manish Kumar singh
YEAR: 2024
PUBLISER : International Journal of Intelligent Systems and Applications in Engineering
Link - Article