基于文本挖掘的建筑工程事故致因与风险管理研究--以辽宁省为例

Research on the Causes and Risk Management of Construction Accidents Based on Text Mining--Through Taking Liaoning Province as an Example

为有效预防和减少建筑工程安全事故的发生,借助ROST CM6内容挖掘系统和UCINET社会网络分析工具,分析2013—2022年辽宁省建筑工程事故调查报告中的事故类型、事故致因、事故地点,以揭示其中的风险规律及相互关联。结果表明,高处坠落在180份事故报告中占比60.56%,是建筑工程中最易发生的事故类型。建筑工程事故最重要的致因是安全意识薄弱、未配戴安全设施,建筑外墙是建筑工程最易发生事故的隐患地点。从已发生的事故入手,在安全风险识别管理中加以关注,制定相应的风险预防措施才能科学有效的规避风险,不断提高建设工程项目的风险管理水平。

In order to effectively prevent and reduce the occurrence of construction project safety accidents,the study uses ROST CM6 content mining system and UCINET social network analysis tool to analyze the accident types,causes and locations of construction project accident investigation reports in Liaoning Province from 2013 to 2022,so as to reveal the risk rules and correlation among them.The results show that falling accounts for 60.56%of 180 accident reports,which is the most common type of accidents in construction engineering.The most important causes of accidents are weak safety awareness and poor safety facilities.Accidents in construction projects mostly occur in building exterior walls.It is suggested to start from the accident,pay attention to the safety risk identification and management,and formulate corresponding risk prevention measures so as to avoid risks scientifically and effectively,and constantly improve the level of risk management of construction projects.