期刊名称：IOP Conference Series: Earth and Environmental Science
印刷版ISSN：1755-1307
电子版ISSN：1755-1315
出版年度：2019
卷号：300
期号：2
页码：1-8
DOI：10.1088/1755-1315/300/2/022130
出版社：IOP Publishing
摘要：In recent years, with the improvement of people's safety awareness and the steady progress of safety production supervision, text classification algorithm based on data mining has been widely applied. At present, for the classification of hidden danger text in coal mine, it mainly relies on manual or machine learning. The efficiency of manual classification is too inefficient to meet the requirements of massive text classification. And the accuracy of machine learning-based classification method is low. In view of the above problems, this paper combines Word2vec and convolutional neural network to achieve accurate classification of hidden danger text in coal mine safety, and achieves great results. The results show that Word2vec can retain the semantic information between contexts. Convolutional neural network can effectively extract the high-level features of local contexts, and the classification effect is more accurate. This method can be implemented in the classification of hidden danger text in coal mine, which has very important practical significance.
