摘要
为探明生产安全事故发生的原因及其内在规律,以期减少生产安全事故或控制事故后果,以文本挖掘任务和社会复杂网络分析方法为切入点,基于R语言对贵州省2017—2023年342份生产安全事故调查报告进行预处理、分词以及特征项降维筛选处理后,获得了28个维度的事故特征项。基于Apriori算法挖掘出事故致因之间的19条强关联规则,基于社会复杂网络分析方法,分别对事故致因进行中心性及核心边缘结构进行分析。结果表明:安全意识淡薄、安全管理不到位、落实主体责任不力、安全教育培训不到位、安全监管不到位、违法违规操作等致因对事故影响较大,在实际安全管控过程中应着重考虑。
To investigate the causes and underlying patterns of production safety accidents with the aim of reducing their occurrence or controlling their consequences,this study employs text mining and social complex network analysis methods.Based on R language,342 production safety accident investigation reports from Guizhou Province(2017—2023)were preprocessed,tokenized,and subjected to dimensionality reduction and feature selection,resulting in 28 accident-related feature items.Using the Apriori algorithm,19 strong association rules among accident causes were mined.Furthermore,centrality and core-periphery structures of the accident causes were analyzed through complex network analysis.The results indicate that factors such as weak safety awareness,inadequate safety management,poor implementation of primary responsibilities,insufficient safety education and training,lax safety supervision,and illegal operations significantly contribute to accidents and should be prioritized in actual safety management practices.
作者
王芳
WANG Fang(Wudang District Emergency Management Bureau,Guiyang Guizhou 550018,China)
出处
《工业安全与环保》
2025年第8期33-38,共6页
Industrial Safety and Environmental Protection
作者简介
王芳(1992-),女,硕士,工程师,主要从事安全、应急管理等方面的研究。E-mail:1348170759@qq.com。
