| 刘鹏,曹新晨,耿念,魏微,孟磊.煤矿标准文件知识图谱构建与应用[J].情报工程,2023,9(5):073-083 |
| 煤矿标准文件知识图谱构建与应用 |
| Construction and Application of Knowledge Graph of Coalmine Standard Documents |
| |
| DOI：10.3772/j.issn.2095-915X.2023.05.006 |
| 中文关键词: 煤矿标准文件；知识图谱；知识抽取；标准信息服务 |
| 英文关键词: Coalmine Standard Documents; Knowledge Graph; Knowledge Extraction; Standard Information Service |
| 基金项目:江苏省安全应急装备技术创新中心（BM2022013）；国家自然科学基金资助项目（71972176）。 |
| 作者 | 单位 | | 刘鹏 | 1. 徐州高新区安全应急装备产业技术研究院 徐州 210000；2. 矿山互联网应用技术国家地方联合工程实验室 徐州 221008 | | 曹新晨 | 3. 中国矿业大学信息与控制工程学院 徐州 221116 | | 耿念 | 3. 中国矿业大学信息与控制工程学院 徐州 221116 | | 魏微 | 3. 中国矿业大学信息与控制工程学院 徐州 221116 | | 孟磊 | 1. 徐州高新区安全应急装备产业技术研究院 徐州 210000；2. 矿山互联网应用技术国家地方联合工程实验室 徐州 221008 |
|
| 摘要点击次数: 734 |
| 全文下载次数: 859 |
| 中文摘要: |
| [目的/意义]为了解决煤矿标准文件知识内容繁多、专业跨度大、组织模式有限，以及缺少系统、高效的煤矿标准文件知识存储利用手段，煤矿标准知识利用率低下等问题，以煤矿相关国家标准和行业标准文件为数据来源，构建
了煤矿标准文件知识图谱。[方法/过程]首先通过分析标准文件特点结构，提出了面向煤矿标准文件的标准文件 - 标准核心要素 - 标准条文的三层知识模型，定义了标准文件层、标准核心要素层以及标准条文层的知识单元结构和实体关
系模式；然后根据设计的图谱模式层，针对半结构化和非结构化数据特点采用基于模板规则和基于深度学习的方法进行知识抽取，并选取图数据库存储方案，实现煤矿标准文件知识存储更新，完成了煤矿标准文件知识图谱构建。[结果/结论]煤矿标准文件知识图谱可实现多种标准可视化方式以增强用户阅读体验，支撑智能检索问答应用帮助从业人员快速检索定位所需标准知识，还能提供标准冲突检测、辅助编写、性能指标比对等多种标准信息服务应用，显著提高了煤矿标准知识利用率，助力煤矿智能化建设。 |
| 英文摘要: |
| [Objective/Significance] Nowadays, many issues exist in application of coalmine standards, including complex knowledge, wide range of specialties, limited organizational patterns, low utilization efficiency and lack of systematic and efficient means for storing and utilizing coalmine standard documents knowledge. To address these issues, a knowledge graph of coalmine standard documents is constructed based on relevant national and industrial standards. [Methods/Processes] Firstly, by analyzing the characteristics and structure of standard documents, a three-layer knowledge model of D-E-P (DocumentsElements-Provisions) for coalmine standards is proposed. The knowledge unit structures and entity relationship patterns of the three layers are defined. Then, according to the schema of the knowledge graph, the knowledge extraction methods based on template rules and deep learning are adopted for semi-structured and non-structured data features, and the graph database was selected to realize the storage and update of coalmine standard knowledge. [Results/Conclusions] The knowledge graph of coalmine standard documents can be used to visualize multiple standards to enhance reading experience, support intelligent retrieval and Q&A applications to help practitioners quickly retrieve and locate required standard knowledge. It can also provide various standard information service applications such as standard conflict detection, auxiliary writing, and performance indicator comparison, significantly improving the utilization efficiency of coalmine standard knowledge and promoting the intelligent construction of coalmines. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
