Public crises are a significant area of research, drawing attention from scholars across diverse fields. Numerous prediction models have been proposed to address real-world challenges associated with crisis events. This study focuses on intelligent prediction models, aiming to predict public crises promptly and mitigate the resultant harm. The paper delves into the application of intelligent computing-based prediction models within the realm of public crisis management. By offering objective data, this research equips decision-makers with valuable information, facilitating the formulation of well-informed plans. The objective is to enhance the decision-making process, introducing scientific rigor and rationality, thereby overcoming the inherent limitations and biases of plans. Experimental results underscore the superiority of the proposed scheme over existing methodologies, showcasing its efficacy in mining crucial information and achieving superior prediction outcomes.