Article’s

The widespread proliferation of misinformation across digital platforms has emerged as a critical challenge to the integrity of public discourse. This paper presents a machine-learning-based Fake News Detection System that classifies news articles as genuine or fabricated using a Random Forest Classifier. Text data undergoes a preprocessing pipeline encompassing tokenization, stop-word removal, and punctuation elimination, followed by feature extraction via Term Frequency-Inverse Document Frequency (TF-IDF) vectorization. An ensemble of decision trees votes collectively to produce the final binary classification label. The model is serialized with Python’s pickle module and deployed through an interactive Streamlit web interface enabling real-time user queries. Experimental evaluation on the Kaggle Fake and Real News benchmark achieves an overall accuracy of 94.5%, with balanced precision and recall across both classes, confirming the viability of ensemble learning for combating online misinformation at scale. Keywords—fake news detection; random forest; TF-IDF; natural language processing; Streamlit; misinformation; ensemble learning.