基于类别关键词权重的煤矿安全隐患分类方法  

Classification Method of Hidden Danger in Coal Mine Safety Based on Weight of Category Keyword

作　　者：林川[1] 武乐飞 戴家佳 LIN Chuan;WU Yuefei;DAI Jiajia(College of Computer Science and Technology,Guizhou University,Guiyang 550025,China;College of Mathematics and Statistics,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学计算机科学与技术学院,贵州贵阳550025 [2]贵州大学数学与统计学院,贵州贵阳550025

出　　处：《贵州大学学报（自然科学版）》2019年第6期53-57,97,共6页Journal of Guizhou University:Natural Sciences

基　　金：贵州省科学技术厅科技支撑计划项目资助(黔科合支撑[2016]2009)

摘　　要：大数据时代,各行各业均产生海量信息,面临大量的信息,如何准确而高效地获取数据中的潜在规律和蕴含价值成为企业信息化的重点。为提升煤矿企业对安全监测数据的理解和监控能力,改善隐患排查治理工作水平,本文提出基于类别关键词权重的短文本分类模型,有效缓解了文本分类中特征稀疏的问题。该方法首先基于朴素贝叶斯算法,对不符合规范的非法数据进行筛选,然后构建基于关键词权重的短文本分类模型,利用中文分词技术、卡方检验方法构建关键词库,最后建立得分模型实现对隐患数据的分类。结果表明,该模型能较为准确地对矿业安全隐患数据进行有效的评级分类,进一步地改善隐患排查和治理的针对性和有效性。In the era of big data,all walks of life generate a large amount of information and produce a large amount of information,and how to accurately and efficiently obtain the potential rules and hidden values in data has become the focus of enterprise informatization.In order to improve the understanding and monitoring ability of coal mine enterprises on safety monitoring data and improve the level of hidden dangers investigation and control,this paper presents a short text classification model based on category keyword weights,which effectively alleviates the problem of sparse features in text classification.Firstly,based on Naive Bayesian algorithm,this method screens the illegal data that does not conform to the norm,and then constructs a short text classification model based on keyword weights.And then the Chinese word segmentation technique and the chi-square test method are used to construct the keyword database.Finally the scoring model is established and the classification of hidden danger data is conducted.The results indicate that the model can accurately classify mining safety hazard data and further improve the pertinence and effectiveness of hazard investigation and management.

关 键 词：权重 短文本分类 煤矿安全隐患 朴素贝叶斯 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TD76[自动化与计算机技术—控制科学与工程]