Unsafe behavior control of miners based on accident reporting excavation

Abstract

Human factor is an important cause of coal mine accidents. In order to study the regularity in human-caused accidents and to improve the target of human-caused management, text data mining method was used to analyze the accident report text data by WordCloud-LDA according to the accident triggering mechanism, and to construct a human-caused factor system for coal mine accidents. On this basis, the human factor matrix set of coal mine accidents is generated, and the correlation analysis of the human factor matrix set is carried out, and the strong association rules between the human factor and the general accident and the roof accident are obtained. The results show that the study of human factors association rules in coal mines under accident report mining can lead to conclusions on the trend among human factors, and provide a basis for human factors control decisions.

Keywords

• incident mining
• text analysis
• wordcloud-lda analysis
• apriori algorithm
• strong association rules