Associative network analysis of inducements for unsafe behaviors based on text-mining method

Abstract

The occurrence of safety accidents in coal mine caused by unsafe behaviors is related to various inducements. This paper aims to explore the correlation between the core inducements and the inducements of unsafe behaviors and their influence on unsafe behaviors. In this paper, 339 coal mine safety accident reports were collected, and 37 unsafe behaviors and their inducement-related words were extracted with the help of text mining method, and the association rules were mined with Apriori algorithm to build the association network of unsafe behaviors and their inducements. Then, the association rules were analyzed. It is show that six types of core inducements, such as insufficient safety awareness, inadequate investigation and management of hidden dangers, and inadequate implementation of the safety production responsibility system, will have a direct impact on unsafe behavior, and six groups of core correlation factor sets will also have an indirect impact. Therefore, coal mining enterprises should focus on the core inducements and their associated factor sets to prevent the occurrence of unsafe behaviors.

Keywords

• coal mine
• text mining
• core inducement
• associative network
• associated factors