Machine Learning-Based Models for Assessing Impacts Before, During and After Hurricane Florence

Abstract

Social media provides an abundant amount of realtime information that can be used before, during, and after extreme weather events. Government officials, emergency managers, and other decision makers can use social media data for decision- making, preparation, and assistance. Social media data and cloud cover temperature as physical sensor data was analyzed in this study using machine learning techniques. Data was collected from Twitter regarding Hurricane Florence from September 11, 2018 through September 20, 2018. Natural language processing models were developed to demonstrate sentiment among the data. Forecasting models for future events were developed for better emergency management during extreme weather events. Different machine learning algorithms and natural language processing techniques were used to examine sentiment classification. The approach is multi-modal, which will help stakeholders develop a more comprehensive understanding of the social impacts of a storm and how to help prepare for future storms. Naive Bayes classifier displayed the highest accuracy for this data. The results demonstrate that machine learning and natural language processing techniques, using Twitter data, are a practical method for sentiment analysis and can be used by decision makers for better emergency management decisions.