基于命名实体识别的水电工程施工安全规范实体识别模型

doi:10.16265/j.cnki.issn1003-3033.2024.09.0008

中国安全科学学报 ›› 2024, Vol. 34 ›› Issue (9): 19-26.doi: 10.16265/j.cnki.issn1003-3033.2024.09.0008

• 安全社会科学与安全管理 • 上一篇下一篇

基于命名实体识别的水电工程施工安全规范实体识别模型

陈述¹^,²(), 张超², 陈云¹^,²^,^**(), 张光飞³, 李智³

¹ 三峡大学水电工程施工与管理湖北省重点实验室,湖北宜昌 443002
² 三峡大学水利与环境学院,湖北宜昌 443002
³ 中国长江三峡集团有限公司,湖北武汉 430010

收稿日期:2024-03-15 修回日期:2024-06-20 出版日期:2024-09-28
通信作者:
^** 陈云(1993—),男,湖北枝江人,博士,副教授,主要从事安全管理研究。E-mail:yunchen@ctgu.edu.cn。
作者简介:
陈述 (1986—),男,湖北英山人,博士,教授,主要从事安全管理研究。E-mail:chenshu@ctgu.edu.cn。
张光飞，高级工程师;
李智，正高级工程师
基金资助:
国家自然科学基金资助(52479127); 国家自然科学基金资助(52079073); 国家自然科学基金资助(52209163)

Model of identifying entities of safety specification for hydropower engineering construction

CHEN Shu¹^,²(), ZHANG Chao², CHEN Yun¹^,²^,^**(), ZHANG Guangfei³, LI Zhi³

¹ Hubei Key Laboratory of Hydropower Engineering Construction and Management, China Three Gorges University, Yichang Hubei 443002, China
² College of Hydraulic & Environmental Engineering, China Three Gorges University, Yichang Hubei 443002, China
³ China Three Gorges Corporation, Wuhan Hubei 430010, China

Received:2024-03-15 Revised:2024-06-20 Published:2024-09-28

摘要/Abstract

摘要：

为准确识别水电工程施工安全规范实体,通过预训练模型中双向编码器表征法(BERT)挖掘文本中丰富的语义信息,利用双向长短期记忆神经网络(BILSTM)提取规范实体语义特征,依靠条件随机场(CRF)分析实体之间的依赖关系,构建水电工程施工安全规范的命名实体识别模型;以《水利水电工程施工安全防护技术规范》(SL714—2015)为例,计算命名实体识别模型精确率。结果表明:BERT-BILSTM-CRF模型准确率为94.35%,相比于3种传统方法,准确率显著提高。研究成果有助于水电工程施工安全规范知识智能管理,为施工安全隐患智能判别提供支撑。

关键词: 命名实体识别, 水电工程施工, 安全规范, 双向编码器表征法(BERT), 双向长短期记忆神经网络(BILSTM), 条件随机场(CRF)

Abstract:

To accurately identify the entities of hydropower engineering construction safety specification, the named entity recognition model of hydropower engineering construction safety specification was constructed. The rich semantic information in the text was mined by the BERT. The semantic features of the specification were extracted by using BILSTM. The dependency relationship between entities was analyzed by relying on CRFs. The Technical Specification for Safety Protection in Construction of Water Conservancy and Hydropower Projects (SL714-2015) was taken as an example to calculate the named entity recognition model accuracy rate. The results show that the accuracy rate of the BERT-BILSTM-CRF model is 94.21%. Compared with the three traditional methods, the accuracy is significantly improved. The research will effectively assist in the intelligent management of safety regulations knowledge for hydropower engineering construction, and provide important support for the intelligent identification of construction safety hazards.

Key words: named entity identification, hydropower engineering construction, safety specification, bidirectional encoder representation from transformers (BERT), bi-directional long and short-term memory neural network (BILSTM), conditional random field (CRF)

中图分类号:

X948

陈述, 张超, 陈云, 张光飞, 李智. 基于命名实体识别的水电工程施工安全规范实体识别模型[J]. 中国安全科学学报, 2024, 34(9): 19-26.

CHEN Shu, ZHANG Chao, CHEN Yun, ZHANG Guangfei, LI Zhi. Model of identifying entities of safety specification for hydropower engineering construction[J]. China Safety Science Journal, 2024, 34(9): 19-26.

图/表 10

图1

图2

表1

表2

表3

表4

图3

图4

表5

表6

参考文献 20

[1]	陈述, 习俊博, 王建平, 等. 水电工程施工安全隐患关联规则挖掘[J]. 中国安全科学学报, 2021, 31(8): 75-82. doi: 10.16265/j.cnki.issn1003-3033.2021.08.011
	CHEN Shu, XI Junbo, WANG Jianping, et al. Mining association rules of near-misses of hydropower projects construction[J]. China Safety Science Journal, 2021, 31(8): 75-82. doi: 10.16265/j.cnki.issn1003-3033.2021.08.011
[2]	樊启祥, 林鹏, 魏鹏程, 等. 水电工程安全事故发生机制与管理对策[J]. 中国安全科学学报, 2019, 29(1): 144-149. doi: 10.16265/j.cnki.issn1003-3033.2019.01.024
	FAN Qixiang, LIN Peng, WEI Pengcheng, et al. Hydropower engineering safety accident occurrence mechanism and management strategies[J]. China Safety Science Journal, 2019, 29(1): 144-149. doi: 10.16265/j.cnki.issn1003-3033.2019.01.024
[3]	CHEN Shu, TIAN Ya, JIN Lianghai, et al. Estimating the frequency of exposure to uncertain hazards: impact of wind conditions on concrete dam construction[J]. Journal of Construction Engineering and Management, 2021, 147(2): 1-13.
[4]	张小衡, 王玲玲. 中文机构名称的识别与分析[J]. 中文信息学报, 1997, 11(4): 21-32.
	ZHANG Xiaoheng, WANG Lingling. Identification and analysis of Chinese institutional names[J]. Journal of Chinese Information Processing, 1997, 11(4): 21-32.
[5]	BIKEL D M, SCHWARTA R, WEISCHEDEL R M. An algorithm that learns what's in a name[J]. Machine Learning Journal Special Issue on Natural Language Learning, 1999, 34(1/2/3): 211-231.
[6]	ZHANG Qiqi, XUE Cong, ZHOU Peng, et al. Named entity recognition for Chinese construction documents based on conditional random field[J]. Frontiers of Engineering Management, 2022, 10(12): 237-249.
[7]	TAEKHYUNG K, SEOKHO C. Accident case retrieval and analyses: using natural language processing in the construction industry[J]. Journal of Construction Engineering and Management, 2019, 145(3): 1-13.
[8]	李明超, 田丹, 沈扬, 等. 融入Attention机制改进Word2vec技术的水利水电工程专业词智能提取与分析方法[J]. 水利学报, 2020, 51(7): 816-826.
	LI Mingchao, TIAN Dan, SHEN Yang, et al. An intelligent extraction and analysis approach of professional technical words for hydraulic engineering by improved Word2vec technology with Attention mechanism[J]. Journal of Hydraulic Engineering, 2020, 51(7): 816-826.
[9]	刘婷, 张社荣, 王超, 等. 水利施工事故文本智能分析的BERT-BiLSTM混合模型[J]. 水力发电学报, 2022, 41(7): 1-12.
	LIU Ting, ZHANG Sherong, WANG Chao, et al. Text intelligent analysis for hydraulic construction accidents based on BERT-BiLSTM hybrid model[J]. Journal of Hydroelectric Engineering, 2022, 41(7): 1-12.
[10]	杨秀璋, 郭明镇, 候红涛, 等. 融合情感词典的改进BILSTM-CNN + Attention情感分类算法[J]. 科学技术与工程, 2022, 22(20): 8761-8770.
	YANG Xiuzhang, GUO Mingzhen, HOU Hongtao, et al. Improved BILSTM-CNN + Attention sentiment classification algorithm fused with sentiment dictionary[J]. Science Technology and Engineering, 2022, 22(20): 8761-8770.
[11]	易明, 李藿然, 刘继月. 基于Glove-BILSTM的在线研讨信息分类模型研究[J]. 情报理论与实践, 2022, 45(9): 173-179,172. doi: 10.16353/j.cnki.1000-7490.2022.09.023
	YI Ming, LI Huoran, LIU Jiyue. Research on online discussion information classification model based on Glove-BILSTM[J]. Information Studies: Theory & Application, 2022, 45(9): 173-179, 172.
[12]	杨秀璋, 彭国军, 李子川, 等. 基于Bert和BiLSTM-CRF的APT攻击实体识别及对齐研究[J]. 通信学报, 2022, 43(6): 58-70. doi: 10.11959/j.issn.1000-436x.2022116
	YANG Xiuzhang, PENG Guojun, LI Zichuan, et al. Research on entity recognition and alignment of APT attack based on Bert and BiLSTM-CRF[J]. Journal on Communications, 2022, 43(6): 58-70. doi: 10.11959/j.issn.1000-436x.2022116
[13]	CHEN Shu, XI Junbo, CHEN Yun, et al. Association mining of near misses in hydropower engineering construction based on convolutional neural network text classification[J]. Computational Intelligence and Neuroscience, 2022, 2022: 1-16.
[14]	赵鹏辉, 李俊杰, 康飞. 河道山地灾害的卷积神经网络快速识别方法[J]. 水利水运工程学报, 2019(2): 65-70.
	ZHAO Penghui, LI Junjie, KANG Fei. Fast recognition method for mountain hazards in river courses based on convolutional neural networks[J]. Hydro-Science and Engineering, 2019, 19(2): 65-70.
[15]	于媛芳, 张勇, 左皓阳, 等. 基于角色信息引导的多轮事件论元抽取[J]. 北京大学学报:自然科学版, 2023, 59(1): 83-91.
	YU Yuanfang, ZHANG Yong, ZUO Haoyang, et al. Multi-turn event argument extraction based on role information guidance[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2023, 59(1): 83-91.
[16]	乔卫亮, 刘阳, 周群, 等. 基于模糊人工神经网络的安全风险评估模型[J]. 安全与环境学报, 2021, 21(4): 1405-1411.
	QIAO Weiliang, LIU Yang, ZHOU Qun, et al. A risk assessment model based on fuzzy artificial neural network[J]. Journal of Safety and Environment, 2021, 21(4): 1405-1411.
[17]	宋敦江, 杨霖, 钟少波. 基于BERT的灾害三元组信息抽取优化研究[J]. 中国安全科学学报, 2022, 32(2): 115-120. doi: 10.16265/j.cnki.issn1003-3033.2022.02.016
	SONG Dunjiang, YANG Lin, ZHONG Shaobo. Research on optimization of disaster triplet information extraction based on BERT[J]. China Safety Science Journal, 2022, 32(2): 115-120. doi: 10.16265/j.cnki.issn1003-3033.2022.02.016
[18]	黄光球, 王瑞泽, 陆秋琴. 基于K-CNN-BiLSTM的关联区域VOCs浓度时空关联预测[J]. 安全与环境学报, 2023, 23(4): 1336-1348.
	HUANG Guangqiu, WANG Ruize, LU Qiuqin. Spatial and temporal correlation prediction of VOCs concentration in associated areas based on K-CNN-BiLSTM[J]. Journal of Safety and Environment, 2023, 23(4): 1336-1348.
[19]	SUBAKTI A, MURFI H, HARIADI N. The performance of BERT as data representation of text clustering[J]. Journal of Big Data, 2022, 9(8): 163-172.
[20]	SL714—2015,水利水电工程施工安全防护设施技术规范[S].
	SL714-2015,Technical specification for safety protection facilities for construction of water and hydropower projects[S].

实体	定义	示例
PLA	施工区域	作业面、进出口
FAC	施工设备	脚手架、钢爬梯
ACT	施工行为	敷设、拆除
PEO	施工人员	员工、作业人员

实体类别	次数
PLA	394
FAC	523
ACT	249
PEO	267

实体类别	P	R	F₁
PLA	91.31	89.55	90.43
FAC	98.12	97.66	97.89
ACT	92.79	91.53	92.16%
PEO	95.18	97.53	96.36

模型	P	R	F₁
BILSTM	78.72	77.07	77.89
LSTM-CRF	86.79	84.51	85.65
BILSTM-CRF	89.52	88.16	88.84
BERT-BILSTM-CRF	94.35	94.07	94.21

模型	P	R	F₁
BERT	67.53	65.05	66.29
BERT-BILSTM	87.65	86.33	86.99
BERT-BILSTM-CRF	94.35	94.07	94.21

基于命名实体识别的水电工程施工安全规范实体识别模型

Model of identifying entities of safety specification for hydropower engineering construction

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 20

相关文章 7

编辑推荐

Metrics

本文评价

模型	识别效果
BILSTM	设有消防/安全/通道,油库/内/道路宜布置成环行道,车道宽应不小于4 m
LSTM-CRF	设有消防/安全通道,油库/内/道路宜布置成环行道,车道宽应不小于4 m
BILSTM-CRF	设有消防/安全通道,油库内/道路宜布置成环行道,车道宽应不小于4 m
BERT-BILSTM-CRF	设有/消防安全通道/,油库内/道路/宜布置成环行道,车道宽应/不小于/4 m

[1]	陈述, 朱丽萍, 陈云, 郑霞忠, 纪勤. 基于复杂网络的水电工程施工安全隐患时序特性*[J]. 中国安全科学学报, 2022, 32(8): 61-66.
[2]	张鹏翔. 多维字符特征表示的铁路设备事故信息抽取方法[J]. 中国安全科学学报, 2022, 32(6): 109-114.
[3]	夏占杰, 张贝克, 高东. 基于数据增强的HSE检查纪要命名实体识别[J]. 中国安全科学学报, 2022, 32(12): 53-62.
[4]	陈述, 刘文濯, 邵波, 习俊博, 王建平. 水电工程施工安全隐患时空分布特征研究[J]. 中国安全科学学报, 2022, 32(11): 90-96.
[5]	陈述, 习俊博, 王建平, 陈云. 水电工程施工安全隐患关联规则挖掘^*[J]. 中国安全科学学报, 2021, 31(8): 75-82.
[6]	郑霞忠，谌巧玲，陈述，胡斌锋. 基于粗糙集的水电工程施工安全评价方法[J]. 中国安全科学学报, 2011, 21(1): 82-.
[7]	孙明义，郝若锦，李忠慧，李进. AQ标准《油漆与粉刷作业安全规范》编制要点[J]. 中国安全科学学报, 2009, 19(8): 172-.