Deep Learning-based Question Answering System for Proactive Disaster Management

As climate change increases the frequency and intensity of natural disasters, proactive disaster management is needed to reduce the damage caused by the natural disasters. Existing reports that record the scale, damage, and response of natural disasters can be used as references for proactive disaster management. However, it is labor-intensive and time-consuming to manually find the necessary information from a number of reports. Thus, this study proposes a natural language processing (NLP)-based question answering system (QA system) for proactive disaster management using the existing reports. This study is focused on paragraphs retrieval, which retrieves paragraphs that have a high similarity to a given question based on the word embedding. The National Hurricane Center's Tropical Cyclone Reports are used to evaluate the proposed method.