Detecting information requirements for crisis communication from social media data: An interactive topic modeling approach
Qing Deng,Yang Gao,Chenyang Wang,Hui Zhang
DOI: https://doi.org/10.1016/j.ijdrr.2020.101692
IF: 4.842
2020-11-01
International Journal of Disaster Risk Reduction
Abstract:<p>Timely and accurate information disclosure is an important step towards open and transparent governance. However, the decision on what to disclose at what time is a difficult issue because information requirements of the public are often vague and time-sensitive, particularly during a major disaster. To facilitate crisis communication, an interactive topic modeling method is proposed to extract the public information requirements and track their evolution from social media. Using a man-made disaster as an example, the proposed method is employed to analyze and present the evolution of public expression using the Chinese microblog data. First, topics from microblog data are discovered through a multi-level topic analysis. The results are interpreted into information requirements related to public demand for investigating the causes and responsible agents behind the Tianjin Port Explosion. Second, a streamgraph method is applied to visualize the topical compositions of contents and their evolution over time. The discovered patterns of information requirements perform retrospective analyses of disasters and provide a basis for concrete recommendations for government disclosing information. Our work has implication on the proactive disclosure mechanism deciding what and when to release formal investigation information to best meet the public's information requirements.</p>
geosciences, multidisciplinary,meteorology & atmospheric sciences,water resources