The frequency and severity of natural disasters around the world have increased due to excessive human interference in the environment. Particularly during major emergencies, social media has significantly changed information-sharing and information-gathering methods. A significant possibility to automatically extract information from a large amount of digital data has also been made possible by sophisticated information extraction models built on machine learning methods. In this paper, different machine learning algorithms are compared based on number of parameters such as precision, recall, F1 score and accuracy. In this paper, we conclude that the accuracy achieved using Support Vector Machine (SVM)linear model is better than all other algorithms.