Natural Disaster
Natural disaster research focuses on improving preparedness, response, and recovery efforts through data analysis and advanced modeling. Current research heavily utilizes machine learning, particularly deep learning architectures like U-Net, BERT, and transformers, to analyze diverse data sources including satellite imagery, social media posts, and weather data for tasks such as damage assessment, needs identification, and misinformation detection. These advancements aim to enhance situational awareness, optimize resource allocation, and ultimately mitigate the societal impacts of natural disasters, contributing significantly to disaster management and risk reduction strategies.
Papers
IKDSumm: Incorporating Key-phrases into BERT for extractive Disaster Tweet Summarization
Piyush Kumar Garg, Roshni Chakraborty, Srishti Gupta, Sourav Kumar Dandapat
PORTRAIT: a hybrid aPproach tO cReate extractive ground-TRuth summAry for dIsaster evenT
Piyush Kumar Garg, Roshni Chakraborty, Sourav Kumar Dandapat