Crowdsourcing is now widely used in disaster management. However, restricted access to social media of vulnerable people in poor areas will pose a risk of selective bias. This study presents a framework to harness the hybrid intelligence of crowd and artificial intelligence (AI) in group decision-making problems to improve the inclusiveness of disaster response decisions. First, crowdsourced data are utilized to extract immediate needs, while an AI algorithm predicts broader population requirements based on area vulnerability features. To improve compassion, the degree of human suffering is estimated by a deep learning method, so as to allocate relief resources to the most urgent victims. To manage the inherent uncertainties in decision-making, a stochastic optimization model is used. Next, a group decision-making method is proposed by incorporating the solutions of the crowd and the AI. A consensus-building approach through mining the maximum consensus sequences is introduced. The framework's efficacy is demonstrated through a real-world case study of the 2021 Henan flood. This study contributes to the field by offering a comprehensive model that enhances real-time disaster response through the combined strengths of crowd and AI intelligence in group decision making.
Tang et al. (Thu,) studied this question.