Performance Evaluation of LLMs with Deep Learning Models for Fake News Detection

Concerns about the integrity of information distribution have grown significantly as a result of the prevalence of fake news on social media and digital platforms. This study presents a comprehensive comparative analysis of the performance of three distinct approaches in fake news detection: standard deep learning models, ChatGPT, and Google Bard. Using the IFND dataset, which contains a varied selection of real-world news items and artificial fake news samples, we assess the efficiency of each technique. We analyze their performance in terms of accuracy, precision, recall, and F1-score through in-depth experimentation. The results offer comprehensive insights into the advantages and disadvantages of each methodology, with deep learning models being particularly good at identifying blatantly obvious fake news, ChatGPT showing promise in handling more complicated situations, and Bard filling the gap. This study highlights the crucial function of large language models while contributing to the ongoing conversation on the identification of fake news. By taking into account the distinctive advantages of each strategy and the potential to increase accuracy in the fight against misinformation, these insights can help direct the development of efficient fake news detection systems.