煤矿事故隐患大数据处理与知识发现分析方法研究

Study on big data processing and knowledge discovery analysis method for safety hazard in coal mine

为解决半结构化或非结构化文本型煤矿隐患数据利用难度大、挖掘深度不够的问题,首先运用六何分析方法对煤矿事故隐患大数据进行内容分析,确定隐患的描述维度及属性类别,实现文本型隐患数据的量化表达;之后根据隐患数据变量特征,采用对数线性模型进行隐患维度间交互的知识发现研究,探索煤矿事故隐患各维度间的交互效应。研究结果表明:基于"六何分析法+对数线性模型"的分析框架能够实现文本型隐患数据的结构化转换,有效揭示煤矿隐患各维度间的交互影响关系,实现隐性知识的显性化。

In order to solve the problem that the semi-structured or unstructured text-based safety hazard data in coal mine is difficult to be utilized and mined deeply,the 5W1 H analysis method was adopted to analyze the content of big data for safety hazard in coal mine. The description dimensions and attribute categories of safety hazard were determined,and the quantitative expression of text-based safety hazard data was realized. Then the log-linear model was used to study the interactive knowledge discovery among the dimensions of safety hazard,and the interaction effect among the dimensions of safety hazard in coal mine was explored. The results showed that the analysis framework based on the ＂ 5W1 H analysis method and log-linear model＂ can achieve the structural transformation of text-based safety hazard data,reveal the interactive influence relationship among the dimensions of safety hazard in coal mine,and realize the changing from implicit knowledge into explicit knowledge.