Leveraging NLP for Fake News Detection: A Comparative Study of ML and DL Models

Today’s modern era is dominated by digital media, in which fake news has become a significant threat to the information authenticity. This paper presents a comprehensive study with comparison of ML and DL approaches. The article presents study on leveraging Natural Language Processing (NLP) for the detection of fake news in digital media and comparing the machine learning (ML) and deep learning (DL) models to consider best performing model in terms of accuracy. The traditional ML algorithms such as Naive Bayes and Support Vector Machines (SVM) alongside DL models like Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNN) are employed. The novelty of this comprehensive study lies in its integrated approach of using multiple ML and DL models to uncover strengths and weaknesses, combined with advanced NLP feature engineering. The results indicate that the CNN model outperforms other models in accuracy. Although SVM shows comparable performance with lower computational requirements.