PREDICTING FAKE INFORMATION IN SOCIAL NETWORK AND AUTHENTICATION USING BLOCK CHAIN

ABSTRACT: Social media's rapid expansion has resulted in a startling increase in the dissemination of false information, endangering public confidence, security, and opinion. This project suggests a hybrid strategy that successfully detects and tracks fake news in social networks by fusing blockchain technology with machine learning. We improve detection accuracy by using a multi-classifier ensemble model that consists of Random Forest, Decision Tree, and Passive Aggressive classifiers. These models are combined in a voting classifier to provide reliable predictions. The system uses blockchain for user authentication and fake news traceability to guarantee transparency and immutability. The implementation uses a two-module architecture: the Participant Node Module, which controls news content, and the Verifier Node Module, which deals with user registration and verification. Keywords — Block Chian, Machine Learning, Random Forest, Decision Tree, Passive Aggressive Classifier.