Explainable misinformation detection across multiple social media platform
Year: 2025
Type: article
Abstract: Social media platforms like Facebook, Social Media, and Instagram have revolutionized communication and interaction, but they have also facilitated the rise of Misinformation, particularly among young users. Misinformation, a form of psychological abuse, can lead to severe mental health issues, including anxiety, depression, and even suicide. Detecting and preventing Misinformation on platforms like Social Media remains a significant challenge due to the dynamic nature of tweets, comments etc..., the use of slang, emojis, and sarcasm, and the anonymity provided to perpetrators. Traditional machine learning and deep learning approaches often struggle to adapt to the continuously evolving content of social media messages. To address these limitations, we propose a novel hybrid approach named DEA-RNN for automated Misinformation detection. DEA-RNN combines Elman-type Recurrent Neural Networks (RNN) with an enhanced Dolphin Echolocation Algorithm (DEA) to optimize network parameters effectively. This approach leverages the strengths of both classifiers and topic modelling, ensuring greater adaptability and improved accuracy in detecting Misinformation instances. Experimental results demonstrate that DEA-RNN outperforms existing models across various evaluation metrics, making it a promising solution for real-time Misinformation detection on Social Media. This study highlights the importance of integrating advanced deep learning models for combating the growing threat of Misinformation on social media platforms
Authors: Dhannina Srinivasa Santha Kumar -, N Ganga Devi -
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W4408971635