Research on emergency decision quality evaluation and optimization basing on public sentiment big data analysis

Abstract

Listen

Translate article

The increasingly frequent occurrence of emergencies has led to significant losses, both human and economic. It is of the utmost importance to scientifically and effectively formulate emergency decision to prevent and reduce these losses. In this study, an evaluation model is developed for evaluating emergency decision quality based on fuzzy comprehensive evaluation method. The proposed model takes into consideration of the impact of public sentiment appeals on emergency decision in emergencies with the purpose of improving emergency decision quality. The “3·21” major explosion accident in Xiangshui County, Jiangsu Province, China, in March of 2019, is selected as a typical case, and the public sentiment big data is obtained from Weibo. The Latent Dirichlet Allocation (LDA) model and the TF-IDF algorithm are utilized to develop an emergency decision quality evaluation index system, from which public sentiment appeals are extracted and quantified. Furthermore, this study explores the changes of public sentiment attitude towards emergency decision through static and dynamic public sentiment analysis. Based on this, the emergency decision quality for the case event is evaluated by the proposed model. The effectiveness of the proposed approach is demonstrated with in-depth discussion and validation through this real case. The results indicate that the emergency decision quality of the Chinese government in this case is fairly good, yet exhibit room for further improvement. Optimization advice for emergency decision in emergencies is proposed accordingly. Empirical evidence suggests that the model proposed in this study can provide theoretical and methodological support for emergency response and government decision-making in the era of big data.