Fake News Detection, Identification and Classification Using Multimodal Approach

Abstract

The influence of social media on individuals and their lives is significant. Numerous news blogs, articles, and other forms of content are regularly disseminated to users through social media platforms. Social media has emerged as a valuable and rapid source of updates. However, the exponential growth in the use of social media and the ease of information sharing have led to a proliferation of dubious and misleading content on the internet. Distinguishing between authentic and fake news has become challenging, as some fabricated news closely resembles genuine news. Consequently, it is crucial to employ automated tools such as machine learning and deep learning techniques for the detection of fake news. In this study, we have employed six models from machine learning and three from deep learning using the ISOT dataset that contains both the fake news and the real ones. For text representation in the machine learning models, we utilized TF-IDF and Bag of Words techniques. The ISOT fake news dataset was utilized for this research. As far as the evaluation is concerned, we used metrics such as recall, precision, accuracy and F1 score. Our proposed GRU model shows a 100% testing accuracy for ISOT dataset, along with 100% recall, precision, and F1 score.