Advancing Fake News Detection Hybrid Deep Lea...

Advancing Fake News Detection Hybrid Deep Learning with FastText with Explainable AI

Author Name : Mr. Muthuvenkatakrishnan R, Selva Kumar A, Nithish Sam Kamalesan D, Mohamed Tawfeeq Ibrahim

ABSTRACT

The pervasive spread of fake news on digital platforms necessitates robust detection mechanisms. This paper presents a system leveraging machine learning for automated fake news classification, integrating PHP for web interface development, SQL for data storage, and Python for algorithm implementation. The system employs Natural Language Processing (NLP) techniques to extract linguistic features, including n-grams and part-of speech tags, alongside sentiment analysis to capture subtle cues indicative of misinformation. Deep learning, specifically a Long Short-Term Memory (LSTM) network, is utilized to model sequential dependencies within news articles, enhancing classification accuracy. Historical news data is incorporated to identify patterns and biases associated with known fake news sources. The system's performance was evaluated using a publicly available dataset, demonstrating a high accuracy rate in distinguishing credible news from fabricated content. The modular design facilitates scalability, enabling deployment on large-scale news platforms. The developed web interface allows users to input news articles for real-time analysis. This research contributes a practical framework for automated fake news detection, addressing the urgent need for reliable information verification in the digital age. Future work will focus on enhancing the model's robustness against adversarial attacks and expanding the system to handle multimodal fake news.

Keywords: Fake News Detection, Deep Learning, Bi-LSTM, CNN, FastText, Explainable AI, SHAP, LIME, SMOTE, Text Classification, Hybrid Model, Natural Language Processing (NLP), Model Interpretability.