FAKE NEWS DETECTION USING DEEP LEARNING MODEL

Abstract

In essence, fake news detection uses machine learning, deep learning, and natural language processing algorithms to recognize and categorize news material as either true or fraudulent. The phrase "fake news" refers to purposefully false or misleading content that is disseminated through social media, certain websites, or messaging apps under the guise of news channels in an effort to sway public opinion, cause confusion, or make money. Examining the textual, linguistic, and contextual elements of news reports and social media articles to ascertain their veracity and authenticity is the primary objective of the fake news detection method. This thesis suggests a deep learning model for efficient fake news detection that combines a gated recurrent unit, convolutional neural networks, and attention mechanisms. The suggested CNN and GRU and Attention model combines the advantages of each element to overcome these drawbacks CNN effectively simulates the sequential nature of text, GRU identifies important phrases and local patterns, and the Attention mechanism draws attention to the most instructive portions of the input. This architecture improves classification's interpretability and accuracy. Experimental results using benchmark false news datasets demonstrate that the proposed model outperforms previous approaches in terms of precision, recall, and F1-score. This gives it a scalable and dependable way to identify bogus news in real time. Demonstrates the efficacy of CNN-GRU-attention model in identifying fake news, providing a robust and comprehensible way to counteract the spread of false information. The creation of even more accurate and dependable systems for identifying fake news has advanced significantly thanks to the synergistic integration of deep learning techniques, supporting ongoing efforts to ensure the accuracy of data transmitted online.