[1]刘 鹏,程浩然,王 莹,等.煤矿工种知识图谱智能问答研究[J].计算机技术与发展,2024,34(03):185-192.[doi:10. 3969 / j. issn. 1673-629X. 2024. 03. 027]
　LIU Peng,CHENG Hao-ran,WANG Ying,et al.Research on Intelligent Question Answering Based on Knowledge Graph of Coalmine Occupation[J].,2024,34(03):185-192.[doi:10. 3969 / j. issn. 1673-629X. 2024. 03. 027]

点击复制

煤矿工种知识图谱智能问答研究()

分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

34

期数:

2024年03期

页码:

185-192

栏目:

人工智能

出版日期:

2024-03-10

文章信息/Info

Title:

Research on Intelligent Question Answering Based on Knowledge Graph of Coalmine Occupation

文章编号:

1673-629X(2024)03-0185-08

作者:

刘 鹏¹; 2 ; 程浩然³ ; 王 莹³ ; 魏 微³ ; 丁恩杰¹; 2

1. 矿山物联网应用技术国家地方联合工程实验室,江苏 徐州 221008;

2. 中国矿业大学 物联网(感知矿山)研究中心,江苏 徐州 221008;

3. 中国矿业大学 信息与控制工程学院,江苏 徐州 221116

Author(s):

LIU Peng1; 2 ; CHENG Hao-ran3 ; WANG Ying3 ; WEI Wei3 ; DING En-jie1; 2

1. The National and Local Joint Engineering Laboratory of Internet Application Technology on Mine,Xuzhou 221008,China;

2. Internet of Things ( Perception Mine) Research Center,China University of Mining and Technology,Xuzhou 221008,China;

3. School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221116,China

关键词:

煤矿工种; 知识图谱; 智能问答; 意图识别; 槽位提取

Keywords:

coalmine occupation; knowledge graph; intelligent question answering; intent recognition; slot extraction

分类号:

TP391. 1

DOI:

10. 3969 / j. issn. 1673-629X. 2024. 03. 027

摘要:

知识图谱是用于表征实体间结构关系的新一代知识库,其通过语义网络描述现实世界事物之间的逻辑关系,而基于知识图谱的智能问答技术也在不断发展,智能问答系统与知识图谱相结合,是对结构化知识的进一步剖析及利用。 该文通过收集煤矿工种专业信息,构建煤矿工种知识图谱,并在此基础上对智能问答技术和系统进行了研究。 在知识图谱构建方面,对工

种专业进行定义,通过 Bert-BiLSTM-CRF 实体识别模型对煤矿工种关键信息进行抽取,再利用图数据库存储三元组工种知识数据得到工种图谱。 在智能问答环节,通过设计问题模板,利用 Bert 模型实现端到端的问句意图识别和槽位提取,并采用 Sentente-Bert 对问句的提及词和知识图谱的候选实体进行链接,继而将问句转化形成图数据库查询语句,从图谱中返回答案。 实验结果表明,构建的煤矿工种知识图谱及智能问答系统,在多个评价指标表现良好,可以满足煤矿工种知识问答需求,为煤矿智能化建设做出了有益探索。

Abstract:

Knowledge graph ( KG) is a new generation of knowledge base used to represent the structural relationship between entities,which describes the logical relationship between things in the real world through the semantic network, and the intelligent questionanswering technology based on knowledge graph is also constantly developing. Knowledge graph based question answering ( KBQA) isthe combination technology of intelligent question answering and knowledge graph,which is a further analysis and utilization of structuredknowledge. In this study,we construct the knowledge graph of coalmine occupation and then study the related KBQA technology andsystem. In terms of knowledge graph construction,the work specialty is defined,the key entity information of coalmine occupation isextracted by Bert-BiLSTM-CRF model,and the work graph is obtained by storing triplet knowledge data of work in graph database. Inthe intelligent question-and-answer session,the question template is designed,Bert model is used to realize the end-

to-end questionintent recognition and slot extraction,and Sentent-BERT is used to link the reference words of the question and the candidate entities ofthe knowledge graph,and then the question is transformed into a graph database query statement,and the answer is returned from thegraph. The experimental results show that the proposed KBQA methodology performs well in several evaluation indicators,which canmeet the requirements of knowledge question answering for coal mines,and make a beneficial exploration for the intelligent constructionof coal mines.

常用功能