A MACHINE LEARNING APPROACH FOR EARLY DETECTION OF SOCIAL MEDIA RUMORS

Social media platforms have become an integral aspect of our lives in the modern day. Social media platforms such as Facebook, Instagram, Twitter, SnapChat, and YouTube are used to connect people and promote companies. Twitter is a massive communication and sharing network where users can express themselves and advertise their companies via 140-character tweets. Each month, around 42 million new Twitter accounts are established. Rumors are easily propagated and spread in crowds, particularly in Online Social Networks (OSNs), owing to their open nature and large user base. Introduced SBR approach for early detection of social media rumors and malicious social bots in this paper. To validate the datasets, to use machine learning algorithms with varied window sizes to pick essential components and local characteristics. Classification was performed using KNN, DT, SVM, NB, and RF classifiers and achieved an accuracy of 97.9%. Experiments using twitter datasets reveal that the proposed model outperforms previous content-based approaches in detecting and verifying early rumors.