Forecasting People’s Needs in Hurricane Events from Social Network

Social networks can serve as a valuable communication channel for calls help, offering assistance, and coordinating rescue activities in disaster. such Twitter allow users to continuously update relevant information, which is especially useful during crisis, where the rapidly changing conditions make it crucial be able access accurate information promptly. media helps those directly affected inform others of on ground real time thus enables workers coordinate their efforts more effectively, better meeting survivors' need. This paper presents new sequence based framework forecasting people's needs disasters using social weather data. It consists two Long Short-Term Memory (LSTM) models, one encodes input sequences other plays conditional decoder that decodes encoded vector forecasts needs. Case studies utilizing data collected Hurricane Sandy 2012, Harvey Irma 2017 were analyzed results compared with obtained statistical language model n-gram an LSTM generative model. Our proposed method forecast successfully than either models. approach shows great promise enhancing disaster management evacuation planning commodity flow management.