Study On Large Group Emergency Decision-Making Method Based On Case And Public Data Analysis

In recent years,serious and mega emergencies have become increasingly frequent,causing huge losses to human lives and social security.The frequent occurrence of crises with severe consequences has brought serious challenges for large group emergency decision-making(EDM).Due to the complexity,variability,uncertainty,and high time pressure of issues,as well as the inadequacy of individual knowledge,experience and relevant information on EDM,it is difficult for decisionmaking experts to grasp the state of affairs and develop emergency response plans,thus they need to quickly and efficiently generate alternative response plans to control emergencies by combining the disposal experience of similar historical cases.Meanwhile,when the crisis breaks out,public behavioral big data on social media platforms reflects their emergency demands,from which group wisdom knowledge can be mined for providing an effective reference for EDM.In addition,since the current large group decision-making(LGDM)is carried out in the framework of social networks,the complex relationships among experts will have important impacts on the decision results,thus how to integrate experts’ complex information and incorporate group complex relationships into the decision framework to assist large group EDM deserves attention.Therefore,focusing on the application of similar case identification and emergency solution generation,public big data and group complex relationships in aiding EDM,we propose a new method for large group EDM based on case and public data analysis in a social network environment.The main research work of this paper is as follows.(1)To address the problem of emergency alternatives generation in large group EDM,a case-based emergency alternatives generation method is proposed,and an approach for two-stage case retrieval considering the feature distribution of case index data and experts’ regret avoidance psychology is constructed,improving the accuracy of similar case screening and case matching efficiency.Firstly,the case index data are processed and interval divided.Then the historical case sets with consistent intervals are identified as alternative cases through the proposed one-stage process.Secondly,considering experts’ emotion of regret avoidance,via the proposed two-stage process,similar cases are retrieved from the identified alternative cases,and their disposal measures are combined to generate alternative response plans for the current event,thus providing informative support for the large group emergency solution selection process of work(3).(2)To address the problems of relying on experts’ subjective determination of attributes and incomplete information for EDM,a group wisdom knowledge mining method based on public data is proposed and applied to the emergency environment.It improves the efficiency of group wisdom knowledge refinement,provides innovative support for capturing the public’s dynamic emergency demands,thus is conducive to enhancing group satisfaction in EDM.Firstly,via big data analysis,improved co-word network analysis and social network analysis,group wisdom knowledge is acquired and visualized from public data on social media platforms.Secondly,considering the influence of public data in the social network and its covered topic information,a value measurement model for group wisdom knowledge is established,which promotes the support of highvalue public data for EDM.Finally,the dynamic decision attribute information generation method is proposed by bringing together experts’knowledge and public group wisdom power,which provides a reference for EDM.(3)Based on work(1)and work(2),the study results are applied to determine the parameters of EDM,containing information on emergency alternatives and decision attributes.And a new method for emergency solution selection based on complex information fusion is proposed for addressing the problem of incomplete consideration of group relationships in LGDM under social network contexts.Through the complex information fusion process,experts’ complex relationships are obtained by synthesizing their features on opinion similarity and trust,and their visualization is achieved by establishing an association network after simulating the clustering process.Based on this,a dynamic decision weight determination method is constructed and a personalized consensus feedback mechanism is designed.The experimental results show that the proposed method can improve the group satisfaction and consensus reaching efficiency,which is conducive to better meet the timely requirements of EDM.