基于文本挖掘的建筑施工坍塌事故致因研究  

Research on causes of collapse accidents in building construction based on text mining

作　　者：田水承 王雪晨[1,2] 范彬彬 TIAN Shuicheng;WANG Xuechen;FAN Binbin(College of Safety Science and Engineering,Xi’an University of Science and Technology,Xi’an 710054,China;Institute of Safety&Emergency Management,Xi’an University of Science and Technology,Xi’an 710054,China)

机构地区：[1]西安科技大学安全科学与工程学院,陕西西安710054 [2]西安科技大学安全与应急管理研究所,陕西西安710054

出　　处：《西安科技大学学报》2022年第5期849-855,共7页Journal of Xi’an University of Science and Technology

基　　金：国家自然科学基金项目(51874237,U1904210);国家社会科学基金项目(20XGL025)。

摘　　要：为明确建筑施工坍塌事故致因及其关系,有效预防事故和减少伤害,选取2014—2020年间国内420例建筑施工坍塌事故调查报告,基于R语言文本挖掘的方法,对所选事故调查报告进行数据清洗、分词、特征选择后,利用WordCloud 2程序包对结果进行可视化展示。运用社会网络分析软件Ucinet对建筑施工坍塌事故致因网络进行中心性、核心-边缘结构及凝聚子群分析。结果表明,TF-IDF算法所得28项事故致因中,11项位于致因网络的核心区域,其中安全意识淡薄、管理混乱、资质不达标、隐患整改不力、监督检查不到位、违章操作等8项致因同样处于词云突出位置,且各致因间联系紧密,应对其高度重视和管控,从而减少建筑施工坍塌事故的发生。In order to clarify the causes of building construction collapse accidents and the relationship between them,and effectively prevent accidents and reduce injuries,the data of 420 domestic building construction collapse accident investigation reports from 2014 to 2020 is taken as the mining language materials.Based on the method of text mining of R language,data cleaning,word segment,and character selection of the accident investigation reports are conducted,and the results are visualized using the WordCloud 2 package.The social network analysis software Ucinet is used to analyze the centrality,core-periphery structure,and cohesive subgroups of the network caused by building collapse accidents.The results show that among the 28 accident causes obtained by the TF-IDF algorithm,11 causes are located in the core area of the cause network,including poor security awareness,chaotic management,substandard qualifications,ineffective rectification of hidden dangers,inadequate supervision and inspection,illegal operations and other 8 ones which are also prominent in the word cloud.And the causes are closely related,which should be highly valueed and controled to reduce the occurrence of building construction collapse accidents.

关 键 词：建筑事故 坍塌致因 文本挖掘 R语言 社会网络分析 

分 类 号：X947[环境科学与工程—安全科学]