Published Paper Details:

The rapid growth of digital media and online social platforms has significantly increased the spread of fake news, which can mislead people and negatively impact public opinion, decision-making, and social harmony. This paper presents a content-based fake news detection system using a hybrid deep learning approach that combines Bidirectional Encoder Representations from Transformers (BERT) and Long Short-Term Memory (LSTM) networks to improve classification accuracy. BERT is utilized to generate contextualized word embeddings by understanding the semantic and syntactic relationships within the text, enabling the model to capture deep linguistic features from news content. The LSTM layer is integrated to effectively learn sequential patterns and long-term dependencies in the data, enhancing the model's ability to distinguish between real and fake information. The dataset used for this study consists of labeled news data from the FakeNewsNet repository, which is preprocessed using standard natural language processing techniques such as tokenization, stopword removal, text cleaning, and normalization. The model is trained and evaluated using performance metrics including accuracy, precision, recall, and F1-score, demonstrating improved results compared to traditional machine learning approaches. The system achieves an overall classification accuracy of approximately 88-90% on the test dataset and provides predictions in under two seconds, confirming its suitability for real-time applications. This work contributes to the development of intelligent systems for detecting misinformation and helps in reducing the harmful effects of fake news in online environments.