Citation

Tan Z, Wang Z, Hu H, Jiang X, Peng S. China Saf. Sci. J. 2019; 29(3): 145-148.

Copyright

(Copyright © 2019, China Occupational Safety and Health Association, Publisher Gai Xue bao)

DOI

10.16265/j.cnki.issn1003-3033.2019.03.024

PMID

unavailable

Abstract

In order to enhance the ability of coal mine safety administrators to understand and process hidden danger data, and improve the work of hidden danger investigation and management, the text clustering method was applied to mining and analyzing the data of historical safety hazards records of coal mining enterprises.The chi-square statistics was used to extract feature words,and the feature words with high degree of correlation with category were used to describe the clustering results. The main types and characteristics of hidden dangers recorded in historical data were studied. The results show that the combination of text clustering and chi-square statistics could identify the types and characteristics of the main hidden dangers recorded in the data on coal mine safety hazards and that the serctor in coal mine responsible for investigation and treatment of hidden dangers should focus on the hidden danger types with large numbers, and formulate corresponding treatment measures according to the characteristics of hidden danger types, so as to improve the pertinence and effectiveness of the investigation and treatment of hidden dangers. © 2019 China Safety Science Journal. All rights reserved.

Language: zh