基于链路预测的安全隐患管理研究*

为探寻安全隐患的发生规律,揭示安全隐患的演化特点,为管理人员提供建议和参考。以潞安集团司马煤业有限公司2009—2015年安全隐患记录为数据源,建立无向加权安全隐患共词网络,利用链路预测技术,寻找安全隐患的内在联系及发展趋势。结果表明:相比于共同邻居、资源分配等基于局部信息的相似性指标,资源分配指标能够更好地预测安全隐患各个关键词之间的变化联系;通过分析关键词共现关系,证实预测的准确性,预测出部分未来存在但当前未产生的节点关系。安全管理人员可基于此方法采取有针对性的预防措施,减少相关隐患的发生。

Research on management of hidden danger based on link prediction

In order to explore the occurrence regularity of hidden danger,reveal the evaluation characteristics of hidden danger,and provide suggestions and reference for the managers,the undirected weighted hidden danger co word networks were established by taking the records of hidden danger in Sima Coal Industry Co.,Ltd.of Lu’an Group from 2009 to 2015 as the data source,then the inherent link and development trend of hidden danger were searched by using the link prediction technology.The results showed that compared with the similarity indicators based on the local information such as the common neighbors and resource allocation,the resource allocation indicator could better predict the change relationship between various keywords of hidden danger.The accuracy of the link predictions was confirmed by analyzing keyword co occurrence relationships,predicting some of the nodal relationships that exist in the future but are not currently generated.On this basis,the safety managers can take targeted prevention measures to reduce the occurrence of relevant hidden danger.