图书情报工作

图书情报工作 >

2026 , Vol. 70 >Issue 12: 104 - 117

DOI: https://doi.org/10.13266/j.issn.0252-3116.2026.12.009

研究论文

中文重名个人名称规范记录实体匹配多元路径研究

  • 陈辰 ,
  • 李佳伦 ,
  • 王雪峰 ,
  • 李春秋 ,
  • 周晓城
展开
  • 1 河北大学管理学院 保定 071100;
    2 邯郸市图书馆 邯郸 056008;
    3 北京师范大学政府管理学院 北京 100875;
    4 香港中文大学 香港 999077
陈辰，副教授，博士，硕士生导师；李佳伦，硕士研究生；王雪峰，副研究馆员，硕士研究生；李春秋，副教授，博士，硕士生导师，通信作者，E-mail：lichunqiu@bnu.edu.cn；周晓城，博士研究生。

收稿日期: 2025-03-31

  修回日期: 2025-08-22

  网络出版日期: 2026-06-09

基金资助

本文系国家社会科学基金项目“名称规范控制向身份管理的转型研究”（项目编号：22BTQ038）研究成果之一。

收起

A Multi-Path Approach to Entity Matching for Chinese Homonymous Personal Name Authority Records

  • Chen Chen ,
  • Li Jialun ,
  • Wang Xuefeng ,
  • Li Chunqiu ,
  • Zhou Xiaocheng
Expand
  • 1 School of Management, Hebei University, Baoding 071100;
    2 Handan City Library, Handan 056008;
    3 School of Government, Beijing Normal University, Beijing 100875;
    4 The Chinese University of Hong Kong, Hong Kong 999077
Chen Chen, Associate Professor, PhD, Master’s Supervisor; Li Jialun, Master’s Candidate; Wang Xuefeng, Associate Research Librarian, Master’s Candidate; Li Chunqiu, Associate Professor, PhD, Master’s Supervisor, Corresponding Author, E-mail: lichunqiu@bnu.edu.cn; Zhou Xiaocheng, Doctoral Candidate.

Received date: 2025-03-31

  Revised date: 2025-08-22

  Online published: 2026-06-09

Supported by

This work is supported by the National Social Science Fund of China project titled “Research on the Transition from Name Authority Control to Identity Management” (Grant No. 22BTQ038).

Fold

摘要

[目的/意义] 研究中文重名规范记录实体匹配的技术路径，综合运用多种算法与策略，旨在为同库或跨库同一人名称规范记录的关联和融合提供借鉴。将大语言模型与规则匹配相结合，提出重名规范记录的实体匹配方案。经数据验证发现该方案可降低中文名称规范数据库数据冗余，有助于实现数据的整合与集成。[方法/过程] 选取具有较高重名率的个人名称规范记录作为分析样本，利用GPT-4o按特定规则从标目附注项和参考数据源等字段提取实体特征，通过中文大语言模型计算嵌入向量，结合最长子字符串匹配、语义距离匹配和层次聚类等技术手段，设计了4种实体匹配方案：提示词+全字段嵌入、单字段嵌入加权、最长子字符串+语义匹配计数、最长子字符串+语义距离加权。[结果/结论] 实验结果表明，最长子字符串+语义匹配计数和单字段嵌入加权方法在F1值上表现突出，召回率和精确率较高，适合图书馆重名规范记录的实体匹配场景。

关键词： 名称规范记录; 重名记录; 实体匹配; 大语言模型; ChatGPT

本文引用格式

陈辰 , 李佳伦 , 王雪峰 , 李春秋 , 周晓城 . 中文重名个人名称规范记录实体匹配多元路径研究[J]. 图书情报工作, 2026 , 70(12) : 104 -117 . DOI: 10.13266/j.issn.0252-3116.2026.12.009

Abstract

[Purpose/Significance] By applying a range of algorithms and strategies, this study explores technical approaches of entity matching for Chinese homonymous authority records. It aims to provide a framework for the association and integration of the same individual within or across databases. By combining large language models with rule-based matching, an entity matching solution for authority records with duplicate names has been proposed. This solution is designed to reduce data redundancy and facilitate the integration of data in Chinese name authority database. [Method/Process] Personal name authority records with high repetition rate were selected as the analysis samples. Using GPT-4o, entity features were extracted from fields such as notes and reference data sources according to specific rules. Embedding vectors were calculated using a Chinese large language model. Four entity matching schemes were designed by combining techniques such as longest common substring matching, semantic distance matching, and hierarchical clustering. These schemes include: prompt + full-field embedding, single-field embedding weighting, longest common substring + semantic match count, longest common substring + semantic distance weighting. [Result/Conclusion] The experimental results show that “longest substring + semantic matching count” and the “single-field embedding weighting” method perform best in terms of F1 score, with high recall and precision. This makes them suitable for entity matching in library homonymous authority records.

Key words： name authority record; homonymous record; entity matching; large language model; ChatGPT

参考文献

[1] 贾君枝, 薛秋红.中文人名名称规范档与维基百科的链接[J]. 图书情报工作, 2015(16): 129-134. (JIA J Z, XUE Q H. The links of Chinese names name authority file with Wikipedia[J]. Library and information service, 2015(16): 129-134.)
[2] 邢鑫. 基于多角度相似性度量的实体匹配方法研究[D]. 北京: 北京交通大学, 2023. (XING X. Research on entity matching based on multi-perspective similarity measurement[D]. Beijing: Beijing Jiaotong University, 2023.)
[3] 维勒. UNIMARC手册：规范格式(第三版)[M]. 中国机读规范格式工作组, 译. 北京: 国家图书馆出版社, 2013. (WILLER M. UNIMARC manual: authorities format (3rd ed.)[M]. China Machine-Readable Format Working Group, trans. Beijing: National Library of China Publishing Press, 2013.)
[4] 鲍彤, 章成志. ChatGPT中文信息抽取能力测评——以三种典型的抽取任务为例[J]. 数据分析与知识发现, 2023, 7(9): 1-11. (BAO T, ZHANG C Z. Extracting Chinese information with ChatGPT: an empirical study by three typical tasks[J]. Data analysis and knowledge discovery, 2023, 7(9): 1-11.)
[5] 梁雯茜. 基于图模型的实体匹配算法研究[D]. 成都: 电子科技大学, 2023. (LIANG W Q. Research on entity matching algorithm based on graph model[D]. Chengdu: University of Electronic Science and Technology of China, 2023.)
[6] 邢昊天. 基于图模型和语义表示的实体链接研究[D]. 南京:东南大学, 2017. (XING H T. Research on entity linking based on graph model and semantic representation[D]. Nanjing: Southeast University, 2017.)
[7] 刘斌. 基于图模型和深度学习的实体链接研究[D]. 武汉: 华中师范大学, 2018. (LIU B. Entity linking based on graph and deep learning[D]. Wuhan: Central China Normal University, 2018.)
[8] 王若霖. 联合强健知识图与属性融合的实体对齐方法研究[D]. 长春:吉林大学, 2024. (WANG R L. Research on entity alignment method based on joint robust knowledge graph and attribute fusion[D]. Changchun: Jilin University, 2024)
[9] 孙辛若. 基于半监督学习的多特征大规模实体分类[D]. 上海: 上海交通大学, 2013. (SUN X R. Multi-feature based semi-supervised learning for large scale entity classification[D]. Shanghai: Shanghai Jiao Tong University, 2013.)
[10] 陈波. 征信系统中实体匹配方法及应用研究[D]. 大连: 大连理工大学, 2010. (CHEN B. The research of entity matching method and application on credit reference system[D]. Dalian: Dalian University of Technology, 2010.)
[11] 于悦. 基于多视角社区发现的实体聚类算法[D]. 北京: 北京化工大学, 2017. (YU Y. Research on entity clustering by multi-views community detection[D]. Beijing: Beijing University of Chemical Technology, 2017.)
[12] 王旭阳, 姜喜秋. 基于上下文信息的中文命名实体消歧方法研究[J]. 计算机应用研究, 2018, 35(4): 1072-1075. (WANG X Y, JIANG X Q. Chinese named entity disambiguation method research based on context information[J]. Application research of computers, 2018, 35(4): 1072-1075.)
[13] 许亮. 基于实体上下文表示学习的实体匹配方法研究[D]. 南京:东南大学, 2020. (XU L. Research on entity matching methods based on entity context representation learning[D]. Nanjing: Southeast University, 2020.)
[14] 吴晓崇, 段跃兴, 张月琴, 等. 基于CNN和深层语义匹配的中文实体链接模型[J]. 计算机工程与科学, 2020, 42(8): 1514-1520. (WU X C, DUAN Y X, ZHANG Y Q, et al. A Chinese entity linking model based on CNN and deep structured semantic model[J]. Computer engineering & science, 2020, 42(8): 1514-1520.)
[15] 张蕊. 基于Bi-LSTM的多领域多范围实体识别研究与实现[D]. 武汉:华中科技大学, 2020. (ZHANG R. Research and implementation of multi-field and multi-range entity recognition based on Bi-LSTM[D]. Wuhan: Huazhong University of Science and Technology, 2020.)
[16] 纪科, 张秀, 马坤, 等.基于关键实体和文本摘要多特征融合的话题匹配算法[J]. 郑州大学学报(工学版), 2024, 45(2): 51-59. (JI K, ZHANG X, MA K, et al. Topic matching algorithm based on multi-feature fusion of key entities and text abstracts[J]. Journal of Zhengzhou university (engineering science), 2024, 45(2): 51-59.)
[17] WANG T S, CHEN X Y, LIN H Y, et al. Match, compare, or select? An investigation of large language models for entity matching[EB/OL]. [2025-09-23]. https://aclanthology.org/2025.coling-main.8.pdf.
[18] 徐正斐, 辛欣. 基于大语言模型的中文实体链接实证研究[J]. 自动化学报, 2025, 51(2): 327-342. (XU Z F, XIN X. An empirical study of Chinese entity linking based on large language model[J]. Acta automatica sinica, 2025, 51(2): 327-342.)
[19] PEETERS R, STEINER A, BIZER C. Entity matching using large language models[EB/OL]. [2025-09-23]. https://arxiv.org/pdf/2310.11244.
[20] PEETERS R, BIZER C. Using ChatGPT for entity matching[EB/OL]. [2025-09-23]. https://doi.org/10.48550/arXiv.2305.03423.
[21] 周烁, 方正, 蒋明鹏, 等. 基于大型语言模型的中文短文本实体链接方法[J]. 指挥信息系统与技术, 2024, 15(6): 41-47. (ZHOU S, FANG Z, JIANG M P, et al. Entity linking method for Chinese short text based on large language model[J]. Command information system and technology, 2024, 15(6): 41-47.)
[22] YANG L, CHENG J, ZHANG F. Advancing cross-lingual entity alignment with large language models: tailored sample segmentation and zero-shot prompts[C]//In Findings of the Association for Computational Linguistics: EMNLP. Miami: Association for Computational Linguistics, 2024: 8122-8138.
[23] 曾泽凡, 胡星辰, 成清, 等. 基于预训练语言模型的知识图谱研究综述[J]. 计算机科学, 2025, 52(1): 1-33. (ZENG Z F, HU X C, CHENG Q, et al. Survey of research on knowledge graph based on pre-trained language models[J]. Computer science, 2025, 52(1): 1-33.)
[24] 李捷佳, 贾君枝. 基于关联数据的作者数据聚合研究[J]. 情报科学, 2019, 37(1): 16-21. (LI J J, JIA J Z. Research on author data aggregation based on linked data[J]. Information science, 2019, 37(1): 16-21.)
[25] 薛秋红, 贾君枝, 刘会洲. 中文名称规范数据与Wikidata语义关联实现[J]. 情报理论与实践, 2019, 42(10): 146-150. (XUE Q H, JIA J Z, LIU H Z. Implementation of semantic association between the Chinese name authority files and Wikidata[J]. Journal of information theory and practice, 2019, 42(10): 146-150.)
[26] 贾君枝, 冯婕. 基于因果链求解算法的人物关系挖掘研究—以Wikidata知识库为例[J]. 情报学报, 2017, 36(3): 221-230. (JIA J Z, FENG J. Research on interpersonal relationships mining based on algorithm of detecting cause-and-effect chains: Wikidata as an example[J]. Journal of information science, 2017, 36(3): 221-230.)
[27] PEETERS R, BIZER C. Dual-objective fine-tuning of BERT for entity matching[C]//Proceedings of the VLDB Endowment. New York: ACM, 2021: 1913-1921.
[28] 郝嘉树. 利用开放语义资源丰富个人名称规范数据——基于FOAF的方案设计[J]. 现代图书情报技术, 2016(2): 75-82. (HAO J S. Enriching personal name authority with open semantic resources: FOAF for schema design [J]. Modern library and information technology, 2016(2): 75-82.)
[29] 王瑞云, 贾君枝. 基于作品关系扩展的中文同名个人规范记录识别与聚簇研究[J]. 图书情报工作, 2017, 61(5): 125-131. (WANG R Y, JIA J Z. Research on the identification and clustering of Chinese same-name personal authority records based on work relationship expansion[J]. Library and information service, 2017, 61(5): 125-131.)
[30] 新浪网. 全国重名率最高的300个名字! 里面一定有你的亲朋好友![EB/OL]. [2025-09-23]. https://k.sina.cn/article_6581118554_18843e65a00100eiew.html. (Xinlang. The 300 Most Common Names Nationwide! There Must Be Your Relatives or Friends in It![EB/OL]. [2025-09-23]. https://k.sina.cn/article_6581118554_18843e65a00100eiew.html.)
[31] 白如江, 陈启明, 张玉洁, 等. 基于ChatGPT+Prompt的专利技术功效实体自动生成研究[J]. 数据分析与知识发现, 2024, 8(4): 14-25. (BAI R J, CHEN Q M, ZHANG Y J, et al. Generating effectiveness entities of patent technology based on ChatGPT+Prompt[J]. Data analysis and knowledge discovery, 2024, 8(4): 14-25.)
[32] 陈辰, 钟雨轩, 李春秋. 名称规范控制和Wikidata身份管理方法的比较研究[J]. 国家图书馆学刊, 2024, 33(5): 94-104. (CHEN C, ZHONG Y X, LI C Q. A comparative study on name authority control and Wikidata identity management methods[J]. Journal of the national library of China, 2024, 33(5): 94-104.)
[33] 段明秀. 层次聚类算法的研究及应用[D]. 长沙: 中南大学, 2010. (DUAN M X. Research and application of hierarchical clustering algorithm[D]. Changsha: Central South University, 2010.)
Options
文章导航

/

〈 〉