Mining Association Rules Of Enterprise Security Hazards Based On Text And Early Warning Method

My country's oil companies have stored a lot of safety management text data after many years of safety management.Because these text data are of various types and are unstructured data,this article aims to find out the management of enterprises hidden in a large number of safety management text data Shortcomings and hidden safety hazards reduce enterprise risks and improve the enterprise's safety management level.This paper uses text mining techniques such as word segmentation and part-of-speech tagging for unstructured text data,combined with association rule algorithm,to construct a mining method for oil company's security risk text data,find out security risks and vulnerabilities,and prepare text mining software in the chemical and oil industry.(1)This paper introduces the Jieba word segmentation and THULAC word segmentation technology into the petroleum industry safety management text data,since there are many professional nouns in the oil industry,in order to reduce the omission of word segmentation,word segmentation discovery and special word merge were adopted.,at the same time,due to the complexity of the text data type,in order to better analyze the data,the word segmentation results were marked,and the equipment was mainly marked seven categories of facilities,institutions,personnel positions,norms and standards ensure the follow-up targeted analysis,at the same time,the safety management text data was extracted according to the labeling categories to form a structured database.Taking two enterprise safety management text data as examples,25916 and 10930 structured databases were formed respectively.(2)For the large text data,in order to effectively mine the association of text data,a text mining model based on Apriori algorithm is established.Combine word segmentation technology to find keywords,use Apriori algorithm,adjust the appropriate support and credibility,find strong association rules,so that you can use strong association rules to analyze the current status of enterprise security management,find management defects and security Hidden dangers,thereby improving safety levels and reducing risks,taking two enterprises as examples,93 and 70 strong association rules were formed respectively.After that,the association rules are displayed in the form of a network diagram.According to the analysis of the generated association rules combined with the network diagram,the company's safety management problems and suggestions are obtained.(3)In order to facilitate the knowledge behind the data and improve the analysis efficiency,this article uses Python language and visualization tool Pyqt algorithm package to develop text mining visualization software.The software is based on the text mining method used in this paper and adds visualization functions.At the same time,it uses a human-computer interaction operation mode,which allows users to adjust in real time during the text mining process and intuitively obtain the association relationship of hidden safety hazards.