A Systematic Approach to Fake News Classification Using the Grapevine SDLC Model

Fake news classification plays a vital role in mitigating misinformation across digital platforms. This research adopts a structured methodology for developing a deep learning-based fake news classifier using the Grapevine Software Development Life Cycle (SDLC) model. The classifier utilizes Bidirectional Encoder Representations from Transformers (BERT) to analyze and categorize news articles as either real or fake. We employ a supervised learning approach incorporating fine-tuned BERT embeddings, dropout layers for regularization, and fully connected layers with the hyperbolic tangent activation function (tanh). Model performance is assessed using accuracy, ROC-AUC score, precision-recall curves, and other statistical metrics, validating the effectiveness of the proposed framework.