主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2023, Vol. 31 ›› Issue (10): 254-265.doi: 10.16381/j.cnki.issn1003-207x.2020.2414

• • 上一篇    

基于质量问题解决的汉语文本数据的因果网络构建方法

王宇彬1,党延忠2(),徐照光2   

  1. 1.中北大学经济与管理学院, 山西 太原 030051
    2.大连理工大学经济管理学院, 辽宁 大连 116024
  • 收稿日期:2020-12-21 修回日期:2021-11-16 出版日期:2023-10-15 发布日期:2023-11-03
  • 通讯作者: 党延忠 E-mail:yzhdang@dlut.edu.cn
  • 基金资助:
    国家自然科学基金资助面上项目(71871041);国家自然科学基金资助青年项目(72001034);山西省高等学校哲学社会科学研究资助项目(2023W085)

A Causal Network Construction Method Based on Chinese Quality Problem Solving Data

Yu-bin WANG1,Yan-zhong DANG2(),Zhao-guang XU2   

  1. 1.School of Economics and Management, North University of China, Taiyuan 030051, China
    2.School of Economics and Management, Dalian University of Technology, Dalian 116024, China
  • Received:2020-12-21 Revised:2021-11-16 Online:2023-10-15 Published:2023-11-03
  • Contact: Yan-zhong DANG E-mail:yzhdang@dlut.edu.cn

摘要:

原因分析是质量问题解决的关键所在。因果关系通常复杂多变,难以仅依靠个人经验准确找到质量问题的根本原因。本文针对实际质量问题解决的汉语文本数据所具有的描述用语不规范、因果结构混杂等特点,提出了适用于这类数据的领域词汇构建方法和因果关系抽取方法,并在此基础上提出了面向质量问题分析的因果网络构建方法。最后,利用某汽车制造企业在白车身冲压生产过程中所记录的质量问题解决数据,构建了具有实际应用意义的因果网络,对本文所提方法的有效性、合理性以及科学性进行了验证。

关键词: 质量管理, QPSD数据, 因果分析, 因果网络, 汽车制造

Abstract:

With the development of economy and technology, people's living standards continue to improve, consumers have higher requirements for the quality and the variety of products. Correspondingly, the quality problems become more and more complicated to solve. Cause analysis is the key to quality problem-solving. However, the causal relationship is usually complex and dynamic. It is challenging to analyze the cause of the problem in-depth only by personal experience. Front-line staffs record the quality problem-solving process, forming the Quality Problem Solving Data (QPS Data) which contains a large amount of knowledge accompanied by the causality of quality problems. However, non-standard records result in non-standard descriptive terms and the mixed causal structure of data. The characteristics mentioned above make it is difficult to apply directly since the data containing much empirical causal knowledge. In order to help improve the effectiveness of cause analysis by utilizing the causal knowledge contained in QPS Data, firstly, relying on the analysis of the characteristics of QPS Data, a bottom-up method is propased to extract domain vocabulary layer by layer from QPS Data, which solves the problem of descriptive terms being non-standardized. Secondly, by analyzing the causal syntactic pattern, a causal relation extraction method based on rule matching is proposed to solve the problem of mixed causal structures. Then, a causal network is constructed to realize the reuse of causal knowledge in QPS Data. Finally, using the quality problem-solving data recorded by an automobile manufacturing enterprise in the process of body-in-white stamping production, a causal network with practical application significance is constructed. And the validity, rationality, and scientificity of the proposed methods are verified. This research can provide extensive and comprehensive support for front-line staff to analyze the cause of the quality problem, help enterprises improve the efficiency of quality problem-solving, and provide enterprises with knowledge management methods for reference. Meanwhile, it contributes to the theories and methods of quality management and knowledge management.

Key words: quality management, quality problem solving data, causal analysis, causal network, automobile manufacturing

中图分类号: