摘要
针对突发事件中群众知识挖掘困难且无法在短时间内收集全面的决策信息等问题,研究构建了一种对重叠网络社区进行公众知识提取的大群体应急决策方法。首先,从社交媒体大数据中提取公众知识,并根据网络用户间的交互数据定义彼此的影响程度,构建用户交互度矩阵,进而通过社区发现算法划分出不同的用户社区,以提取各类公众的差异化决策意见,确定方案的属性权重;其次,考虑到重叠社区发现算法(Community Overlap Propagation Algorithm, COPRA)在庞大、稀疏的网络社区中无法完成快速划分,对COPRA算法的标签更新步骤进行了优化,并将其拓展至带权图;再次,以决策专家信任关系为基础构建社会网络,计算专家个体权重,通过融合专家偏好与方案属性权重得到方案得分及其排序结果;最后,将重叠社区群智融合应急决策方法应用于公众安全事件案例中,并与其他文献的方法进行对比分析,验证了该方法的合理性和有效性。
This paper addresses the challenges of extracting knowledge from the public during emergencies and the difficulty of gathering comprehensive decision-making information within a limited timeframe.It proposes a large-group emergency decision-making method that leverages the wisdom of the group from overlapping network communities.Firstly,public wisdom is extracted from social media big data.The degree of influence among users is defined based on their interaction data,and a user interaction matrix is constructed.A community discovery algorithm is then applied to identify distinct user communities.This approach enables the extraction of differentiated decision-making opinions from various public groups,which are subsequently used to determine the attribute weights of the proposed solutions.Secondly,recognizing that the COPRA algorithm struggles with efficient community division in large and sparse networks,the label update step of the COPRA algorithm is optimized.This optimization allows the algorithm to focus on prominent nodes within the network,rather than peripheral nodes with less influence on the community.The goal of this enhancement is to improve the algorithm s performance in large-scale,sparse networks.Additionally,the algorithm is extended to band-weighted graphs,incorporating supplementary information to further refine community segmentation.Following this,a social network is constructed based on the trust relationships among decision-making experts.The weights of individual experts are calculated,and the fusion of the experts preferences with the scheme attribute weights is used to derive the scheme scores.These scores are aggregated to obtain the overall program ranking results.Finally,the method proposed in this paper is applied to public safety event cases,and its performance is compared and analyzed against methods from other literature.This comparison validates the reasonableness and effectiveness of the proposed method.
作者
陈兆芳
黄鹏城
黄文翰
CHEN Zhaofang;HUANG Pengcheng;HUANG Wenhan(School of Management,Fujian University of Technology,Fuzhou 350118,China;School of Internet Economics and Business,Fujian University of Technology,Fuzhou 350014,China;Fujian Special Equipment Inspection&Research Institute,Fuzhou 350008,China)
出处
《安全与环境学报》
北大核心
2025年第6期2279-2290,共12页
Journal of Safety and Environment
基金
福建省科协科技创新智库课题(FJKX2024XKB014)。
关键词
公共安全
应急决策
重叠社区发现算法
群智提取
社会网络
public safety
emergency decision making
community overlap propagation algorithm
crowd intelligence extraction
social networks
