基于文本挖掘的电力行业触电人身伤亡事故致因分析研究OA

Electric shock accident is the most common type of accident in the construction process of the power industry except for high fall.Identifying the influencing factors leading to electric shock acci-dents is the premise of taking targeted safety management measures.Based on text mining technology and R language,word segmentation and word frequency analysis were carried out on 80 reports of elec-tric shock casualties in power industry construction.By constructing social network and semantic net-work analysis diagrams and conducting centrality analysis,58 influencing factors were divided into four levels:key factors,important factors,secondary factors and general factors.The extracted factors were sentenced according to the related conjunctions,and the influencing factors with similar main contents were merged and summarized or deleted,so as to obtain the electric shock accident cause factor library in power industry.By combining the cause factor database of electric shock accidents with the results of centrality analysis,the most critical factors of electric shock accidents are obtained,which are unautho-rised work(A3),inadequate or unimplemented occupational safety and health responsibility(D1)and defective protective devices and facilities(B1),should be highly valued and focused on control.