Data Analytics for Disaster Management Response using Sentiments

Social media has a very significant influence on numerous fields in the modern world. These social media platforms generate big data that can be used to analyze an event. The impacts of catastrophic events like natural disasters could be efficiently managed by the analysis of this huge amount of data. Information-sharing platforms like Twitter have proven extremely helpful in gathering information to offer data linked to relief efforts run by NGOs or volunteers. The extraction of tweets from the Twitter platform using the Twitter application programming interface, their classification into “need of resources,” “availability of resources,” and irrelevant tweets with the help of deep learning, clustering, and various natural language processing techniques like BERT, distilBERT, RF, LR, NB, and SVM classification models, and the comparison of their results are aimed to be conducted in this research.