Combating Digital Deception: Machine Learning Approaches for Detecting Political Misinformation and Clickbait on Social Media

This paper presents machine learning-based strategies for mitigating the problem of misinformation in social media, with a special focus on political misinformation detection and clickbait detection. The following research work attempts to address the pressing issue of the rapid spread of false or misleading information by making use of deep recurrent neural networks (RNN s) combined with semi-supervised learning methodologies. For political misinformation detection, a hybrid model integrating Long Short-Term Memory (LSTM) networks with fully connected layers was developed, achieving 94.22 % accuracy on the CIC Truth Seeker dataset. Clickbait detection was made using a sequential architecture that combines LSTM and Gated Recurrent Units (GRU) layers with Simple RNN layers to achieve an accuracy of 96% on test data. This study underlines the effectiveness of different machine learning techniques in battling misinformation that would contribute to online dignity in political discourse and ensure better public communication in a digital world.