Performance Evaluation of Diverse ML Techniques in Disasters Twitter classification

Social media has grown to be an essential and significant aspect of our lives in the modern day. Due to the widespread usage of laptops, mobile phones, and other devices, individuals find it simpler to communicate with one another and to post their opinions, concerns, and other information on social media sites like Facebook, Instagram, and tweeter. Several government and commercial organizations use these social media platforms to gather text and assess people’s sentiments during times of crisis. Tracking the tweets relating to the accident is too tough for them to do. To categories tweets as “disaster” or “non-disaster,” the paper implemented three models, the Naïve-Bayes, KNN and support vector machine classification models. The Naïve-Bayes algorithm shows the highest accuracy of 86.591.