A Novel Temporal Footprints-Based Framework for Fake News Detection

With the evolution of social media platforms, the detection of fake news and misinformation is gaining popularity. Social media platforms are the fastest source of fake news propagation, whereas online news websites contribute to dissemination. In recent studies, the temporal features in text documents have gained valuable consideration from the natural language processing (NLP) research community. This study investigates the importance of temporal features in text documents for detecting fake news. Later, the temporal features are combined with the textual features to increase classifier performance. This research study uses Random Forest (RF) and Bi-LSTM techniques to classify fake news based on temporal features and textual features. A publicly available dataset was used to train and test the model. The experimental results demonstrated that the proposed method achieved 99% accuracy by combining temporal and textual features in fake news detection.