FACTUALITY: BERT TRANSFORMER FOR SEMANTIC ANALYSIS AND FAKE NEWS CLASSIFICATION
Keywords:
BERT, Fake news detection, Misinformation, Transformer models, Semantic analysis, Contextual embeddings, NLP, Article classification, Deep learning, Bidirectional language modelAbstract
The Advancements in technology have led to difficulty in finding authentic news and fake news. The BERT-based transformer model enables the identification of fake news and authentic news across the internet. BERT enhances the model through complex operations, which it can perform, such as deep semantic analysis, pattern matching techniques and rich contextual understanding. The model is trained with authentic news and fake news to understand the patterns between real news and fake news to identify warning signs of misinformation, including internal contradictions, peculiar linguistic patterns, and structural inconsistencies.
Leveraging BERT in this model enabled the ability to analyse patterns and detect fake news on various platforms such as healthcare, finance, media, articles, websites. Another major advantage of BERT is that it enhances the accuracy and outperforms the speed of traditional machine learning techniques. Therefore offering a optimized solution toward less misinformation online
