Article(id=1269679023706772060, tenantId=1146029695717560320, journalId=1269656373470969926, issueId=1269678996485734867, articleNumber=null, orderNo=null, doi=10.3969/j.issn.1008-0821.2026.03.003, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1765382400000, receivedDateStr=2025-12-11, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1780644668746, onlineDateStr=2026-06-05, pubDate=1772294400000, pubDateStr=2026-03-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1780644668746, onlineIssueDateStr=2026-06-05, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1780644668746, creator=13701087609, updateTime=1780644668746, updator=13701087609, issue=Issue{id=1269678996485734867, tenantId=1146029695717560320, journalId=1269656373470969926, year='2026', volume='46', issue='3', pageStart='3', pageEnd='183', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1780644662255, creator=13701087609, updateTime=1780644725097, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1269679260173234368, tenantId=1146029695717560320, journalId=1269656373470969926, issueId=1269678996485734867, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1269679260173234369, tenantId=1146029695717560320, journalId=1269656373470969926, issueId=1269678996485734867, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=30, endPage=43, ext={EN=ArticleExt(id=1269679024524661343, articleId=1269679023706772060, tenantId=1146029695717560320, journalId=1269656373470969926, language=EN, title=Research on Entity Alignment in Chinese Materia Medica Knowledge Graphs for Knowledge Fusion, columnId=1269679024180728413, journalTitle=Journal of Modern Information, columnName=DATA INTELLIGENCE and KNOWLEDGE SERVICE, runingTitle=null, highlight=null, articleAbstract=
Purpose/Significance

The digital transformation of Chinese Materia Medica(CMM) classics is critical for bridging ancient pharmacological wisdom with modern drug discovery. However, existing Knowledge Graphs(KGs) for CMM are often constructed in isolation, resulting in fragmented information silos that hinder global data interoperability. While Entity Alignment(EA) has become a focal point in the international Semantic Web community, specific research targeting the alignment of ancient CMM literature remains a significant gap. Moreover, current state-of-the-art models—primarily designed for modern, high-resource languages—struggle to address the unique challenges of ancient Chinese texts. These challenges include severe structural heterogeneity caused by disparate historical writing styles, high terminological ambiguity where distinct medical concepts share similar characters, and a critical scarcity of high-quality annotated data⁃sets. This study aims to fill this gap by proposing a domain-specific deep learning framework designed to automate the fusion of multi-source historical medical knowledge.

Method/Process

To overcome these barriers, this paper proposed the Generative Adversarial Fuzzy-boundary Learning(GAFL-Align) model. The study utilized two representative classics spanning different historical eras: Shennong Bencao Jing and Tangye Bencao. After data cleaning, the datasets comprised 3 771 and 3 910 normalized entities, respectively, focusing on core categories such as herbs, symptoms, and diseases. The technical architecture integrated BERT for deep semantic encoding with Graph Attention Networks(GAT) to capture topological structures. To handle distribution shifts across heterogeneous texts, the model employed a Generative Adversarial Network(GAN) for domain adaptation, mapping entities into a unified feature space. Furthermore, a novel fuzzy boundary negative sampling strategy was developed to distinguish “hard negatives”—terms with high lexical similarity but distinct medical meanings. To address data scarcity, an iterative self-training mechanism with confidence-aware filtering was implemented to augment the training set from a limited number of expert-annotated seed pairs.

Result/Conclusion

Experimental results indicated that GAFL-Align achieved a Hits@1 score of 83.59%, significantly outperforming nine baselines, including translation-based models, GNN variants, and Large Language Models(LLMs)-augmented approaches like ChatEA. The model successfully constructed a fused KG containing 6 826 entities, effectively merging heterogeneous data while preserving unique source-specific attributes. These findings demonstrate that combining adversarial domain adaptation with fine-grained semantic differentiation offers a superior solution for low-resource historical knowledge fusion compared to generic LLMs. Ultimately, this research provides a robust technical foundation for the intelligent organization of CMM heritage, offering significant implications for digital humanities and the global standardization of traditional medicine data.

, correspAuthors=He Li, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=He Li, Wenshi Shao, Jiayu Liu, Jinyuan Zhang, Wang Shen, Guimin Wang), CN=ArticleExt(id=1269679029268419187, articleId=1269679023706772060, tenantId=1146029695717560320, journalId=1269656373470969926, language=CN, title=面向知识融合的本草典籍知识图谱实体对齐研究, columnId=1269679024457544608, journalTitle=现代情报, columnName=数据智能与知识服务, runingTitle=null, highlight=null, articleAbstract=
目的/意义

针对本草典籍知识图谱实体对齐任务中图谱异构、术语易混淆及高质量标注稀缺等挑战,提出融合生成对抗网络与模糊语义辨识的实体对齐模型GAFL-Align,旨在实现多源知识自动化融合。

方法/过程

该模型通过BERT与图注意力网络融合实体语义与拓扑结构,利用生成对抗网络进行领域自适应以消除异构引发的特征分布差异,采用模糊边界负采样策略强化对易混淆术语的细粒度辨识,并结合迭代自训练机制利用高置信度结果扩充样本,有效降低对人工标注的依赖。

结果/结论

实验表明,该模型在自建数据集上的核心指标均优于基线方法。在此基础上构建的多源融合图谱实现了典籍间知识的互补与增值,为本草典籍知识自动化融合提供了有力的技术支撑。

, correspAuthors=李贺, authorNote=null, correspAuthorsNote=
李贺(1964-),女,教授,博士,博士生导师,研究方向:信息行为分析、知识管理。
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=dVrXY9qa9nT8hjiIYnMPfQ==, magXml=NWAffiJMn+/37s2F2rbJmw==, pdfUrl=null, pdf=T9TvsRVikT1Vi4L1xImURg==, pdfFileSize=7170824, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=bkYTxRUx98ANw1fJcXBQVA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=sHZTSo2Mk8tEePjCIK96uw==, mapNumber=null, authorCompany=null, fund=null, authors=

邵文诗(1999-),女,硕士研究生,研究方向:中医古籍数字人文

刘嘉宇(1995-),男,助理研究员,博士,研究方向:数据挖掘

张津源(2001-),女,博士研究生,研究方向:中医古籍数字人文

沈旺(1983-),女,教授,博士,博士生导师,研究方向:用户信息行为

王桂敏(1976-),女,研究馆员,硕士,研究方向:科技信息服务。

, authorsList=李贺, 邵文诗, 刘嘉宇, 张津源, 沈旺, 王桂敏)}, authors=[Author(id=1269679030447018621, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1269679030816117376, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679030447018621, language=EN, stringName=He Li, firstName=He, middleName=null, lastName=Li, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1School of Business and Management,Jilin University,Changchun130012,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1269679030912586369, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679030447018621, language=CN, stringName=李贺, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, *, address=1吉林大学商学与管理学院,吉林长春130012, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1269679029989839477, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=1, ext=[AuthorCompanyExt(id=1269679030006616694, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1School of Business and Management,Jilin University,Changchun130012,China), AuthorCompanyExt(id=1269679030019199607, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1吉林大学商学与管理学院,吉林长春130012)])]), Author(id=1269679031239742083, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1269679031352988293, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679031239742083, language=EN, stringName=Wenshi Shao, firstName=Wenshi, middleName=null, lastName=Shao, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1School of Business and Management,Jilin University,Changchun130012,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1269679031726281350, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679031239742083, language=CN, stringName=邵文诗, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1吉林大学商学与管理学院,吉林长春130012, bio={"content":"

邵文诗(1999-),女,硕士研究生,研究方向:中医古籍数字人文

"}, bioImg=null, bioContent=

邵文诗(1999-),女,硕士研究生,研究方向:中医古籍数字人文

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1269679029989839477, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=1, ext=[AuthorCompanyExt(id=1269679030006616694, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1School of Business and Management,Jilin University,Changchun130012,China), AuthorCompanyExt(id=1269679030019199607, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1吉林大学商学与管理学院,吉林长春130012)])]), Author(id=1269679032074408584, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1269679032191849098, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679032074408584, language=EN, stringName=Jiayu Liu, firstName=Jiayu, middleName=null, lastName=Liu, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1School of Business and Management,Jilin University,Changchun130012,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1269679032498033291, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679032074408584, language=CN, stringName=刘嘉宇, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1吉林大学商学与管理学院,吉林长春130012, bio={"content":"

刘嘉宇(1995-),男,助理研究员,博士,研究方向:数据挖掘

"}, bioImg=null, bioContent=

刘嘉宇(1995-),男,助理研究员,博士,研究方向:数据挖掘

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1269679029989839477, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=1, ext=[AuthorCompanyExt(id=1269679030006616694, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1School of Business and Management,Jilin University,Changchun130012,China), AuthorCompanyExt(id=1269679030019199607, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1吉林大学商学与管理学院,吉林长春130012)])]), Author(id=1269679032602890893, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1269679032980378255, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679032602890893, language=EN, stringName=Jinyuan Zhang, firstName=Jinyuan, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1School of Business and Management,Jilin University,Changchun130012,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1269679033349477008, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679032602890893, language=CN, stringName=张津源, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1吉林大学商学与管理学院,吉林长春130012, bio={"content":"

张津源(2001-),女,博士研究生,研究方向:中医古籍数字人文

"}, bioImg=null, bioContent=

张津源(2001-),女,博士研究生,研究方向:中医古籍数字人文

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1269679029989839477, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=1, ext=[AuthorCompanyExt(id=1269679030006616694, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1School of Business and Management,Jilin University,Changchun130012,China), AuthorCompanyExt(id=1269679030019199607, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1吉林大学商学与管理学院,吉林长春130012)])]), Author(id=1269679033450140306, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1269679033827627668, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679033450140306, language=EN, stringName=Wang Shen, firstName=Wang, middleName=null, lastName=Shen, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1School of Business and Management,Jilin University,Changchun130012,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1269679034175754901, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679033450140306, language=CN, stringName=沈旺, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1吉林大学商学与管理学院,吉林长春130012, bio={"content":"

沈旺(1983-),女,教授,博士,博士生导师,研究方向:用户信息行为

"}, bioImg=null, bioContent=

沈旺(1983-),女,教授,博士,博士生导师,研究方向:用户信息行为

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1269679029989839477, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=1, ext=[AuthorCompanyExt(id=1269679030006616694, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1School of Business and Management,Jilin University,Changchun130012,China), AuthorCompanyExt(id=1269679030019199607, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1吉林大学商学与管理学院,吉林长春130012)])]), Author(id=1269679034276418199, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1269679034611962521, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679034276418199, language=EN, stringName=Guimin Wang, firstName=Guimin, middleName=null, lastName=Wang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Jilin Provincial Library,Changchun130022,China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1269679034704237210, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, authorId=1269679034276418199, language=CN, stringName=王桂敏, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2吉林省图书馆,吉林长春130022, bio={"content":"

王桂敏(1976-),女,研究馆员,硕士,研究方向:科技信息服务。

"}, bioImg=null, bioContent=

王桂敏(1976-),女,研究馆员,硕士,研究方向:科技信息服务。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1269679030111474296, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=2, ext=[AuthorCompanyExt(id=1269679030124057209, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679030111474296, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Jilin Provincial Library,Changchun130022,China), AuthorCompanyExt(id=1269679030136640122, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679030111474296, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2吉林省图书馆,吉林长春130022)])])], keywords=[Keyword(id=1269679035081724571, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, orderNo=1, keyword=knowledge fusion), Keyword(id=1269679035429851804, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, orderNo=2, keyword=entity alignment), Keyword(id=1269679035513737885, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, orderNo=3, keyword=Chinese Materia Medica classics), Keyword(id=1269679035870253726, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, orderNo=4, keyword=knowledge graph), Keyword(id=1269679036293878431, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, orderNo=5, keyword=deep learning), Keyword(id=1269679036688143008, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, orderNo=1, keyword=知识融合), Keyword(id=1269679036809777825, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, orderNo=2, keyword=实体对齐), Keyword(id=1269679037174682274, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, orderNo=3, keyword=本草典籍), Keyword(id=1269679037266956963, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, orderNo=4, keyword=知识图谱), Keyword(id=1269679037522809508, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, orderNo=5, keyword=深度学习)], refs=[Reference(id=1269679040077140669, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=66, issue=22, pageStart=4, pageEnd=15, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=沈旺, 刘嘉宇, 李贺, journalName=图书情报工作, refType=null, unstructuredReference=沈旺,刘嘉宇,李贺,.学科领域生命周期视域下国内中医古籍数字化研究进展[J].图书情报工作202266(22):4-15., articleTitle=学科领域生命周期视域下国内中医古籍数字化研究进展, refAbstract=null), Reference(id=1269679040181998270, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2024, volume=48, issue=1, pageStart=228, pageEnd=233, url=null, language=null, rfNumber=2, rfOrder=1, authorNames=赵辉, 张华敏, 张磊, journalName=中国中医药图书情报杂志, refType=null, unstructuredReference=赵辉,张华敏,张磊,.中医古籍文献研究方法述要[J].中国中医药图书情报杂志202448(1):228-233., articleTitle=中医古籍文献研究方法述要, refAbstract=null), Reference(id=1269679040270078655, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2024, volume=21, issue=20, pageStart=62, pageEnd=66, url=null, language=null, rfNumber=3, rfOrder=2, authorNames=谢云霏, 贾李蓉, journalName=中国医药导报, refType=null, unstructuredReference=谢云霏,贾李蓉.中医药知识图谱构建技术及应用的研究进展[J].中国医药导报202421(20):62-66., articleTitle=中医药知识图谱构建技术及应用的研究进展, refAbstract=null), Reference(id=1269679040341381824, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=38, issue=3, pageStart=272, pageEnd=278, url=null, language=null, rfNumber=4, rfOrder=3, authorNames=王松, 李正钧, 杨涛, journalName=南京中医药大学学报, refType=null, unstructuredReference=王松,李正钧,杨涛,.中医药知识图谱研究现状及发展趋势[J].南京中医药大学学报202238(3):272-278., articleTitle=中医药知识图谱研究现状及发展趋势, refAbstract=null), Reference(id=1269679040429462209, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=10, issue=9, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=4, authorNames=Weng H, Chen J L, Ou A H, journalName=JMIR Medical Informatics, refType=null, unstructuredReference=Weng HChen J LOu A Het al.Leveraging Representation Learning for the Construction and Application of a Knowledge Graph for Traditional Chinese Medicine:Framework Development Study[J].JMIR Medical Informatics202210(9):e38414., articleTitle=Leveraging Representation Learning for the Construction and Application of a Knowledge Graph for Traditional Chinese Medicine:Framework Development Study, refAbstract=null), Reference(id=1269679040542708418, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2020, volume=13, issue=12, pageStart=2326, pageEnd=2340, url=null, language=null, rfNumber=6, rfOrder=5, authorNames=Sun Z Q, Zhang Q H, Hu W, journalName=Proceedings of the VLDB Endowment, refType=null, unstructuredReference=Sun Z QZhang Q HHu Wet al.A Benchmarking Study of Embedding-Based Entity Alignment for Knowledge Graphs[J].Proceedings of the VLDB Endowment202013(12):2326-2340., articleTitle=A Benchmarking Study of Embedding-Based Entity Alignment for Knowledge Graphs, refAbstract=null), Reference(id=1269679040664343235, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2023, volume=7, issue=9, pageStart=146, pageEnd=158, url=null, language=null, rfNumber=7, rfOrder=6, authorNames=翟东升, 娄莹, 阚慧敏, journalName=数据分析与知识发现, refType=null, unstructuredReference=翟东升,娄莹,阚慧敏,.基于多源异构数据的中医药知识图谱构建与应用研究[J].数据分析与知识发现20237(9):146-158., articleTitle=基于多源异构数据的中医药知识图谱构建与应用研究, refAbstract=null), Reference(id=1269679040735646404, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=1095, pageEnd=1104, url=null, language=null, rfNumber=8, rfOrder=7, authorNames=Mao X, Wang W T, Xu H M, journalName=null, refType=null, unstructuredReference=Mao XWang W TXu H Met al.Relational Reflection Entity Alignment[C]//Proceedings of the 29th ACM International Conference on Information and Knowledge Management.New York:Association for Computing Machinery,2020:1095-1104., articleTitle=Relational Reflection Entity Alignment, refAbstract=null), Reference(id=1269679040802755269, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=3174, pageEnd=3180, url=null, language=null, rfNumber=9, rfOrder=8, authorNames=Tang X B, Zhang J, Chen B, journalName=null, refType=null, unstructuredReference=Tang X BZhang JChen Bet al.BERT-INT:A BERT-Based Interaction Model for Knowledge Graph Alignment[C]//Bessiere C.Proceedings of the 29th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2020:3174-3180., articleTitle=BERT-INT:A BERT-Based Interaction Model for Knowledge Graph Alignment, refAbstract=null), Reference(id=1269679040874058438, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2016, volume=37, issue=4, pageStart=8, pageEnd=13, url=null, language=null, rfNumber=10, rfOrder=9, authorNames=阮彤, 孙程琳, 王昊奋, journalName=医学信息学杂志, refType=null, unstructuredReference=阮彤,孙程琳,王昊奋,.中医药知识图谱构建与应用[J].医学信息学杂志201637(4):8-13., articleTitle=中医药知识图谱构建与应用, refAbstract=null), Reference(id=1269679040957944519, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=1511, pageEnd=1517, url=null, language=null, rfNumber=11, rfOrder=10, authorNames=Chen M H, Tian Y T, Yang M H, journalName=null, refType=null, unstructuredReference=Chen M HTian Y TYang M Het al.Multilingual Know⁃ledge Graph Embeddings for Cross-Lingual Knowledge Alignment[C]//Sierra C.Proceedings of the 26th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2017:1511-1517., articleTitle=Multilingual Know⁃ledge Graph Embeddings for Cross-Lingual Knowledge Alignment, refAbstract=null), Reference(id=1269679041046024904, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2017, volume=null, issue=null, pageStart=628, pageEnd=644, url=null, language=null, rfNumber=12, rfOrder=11, authorNames=Sun Z Q, Hu W, Li C K, journalName=null, refType=null, unstructuredReference=Sun Z QHu WLi C K.Cross-Lingual Entity Alignment via Joint Attribute-Preserving Embedding[C]//DAmato C,Fernandez M,Tamma V,et al.Semantic Web-ISWC 2017.Cham:Springer,2017:628-644., articleTitle=Cross-Lingual Entity Alignment via Joint Attribute-Preserving Embedding, refAbstract=null), Reference(id=1269679041134105289, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=5429, pageEnd=5435, url=null, language=null, rfNumber=13, rfOrder=12, authorNames=Zhang Q H, Sun Z Q, Hu W, journalName=null, refType=null, unstructuredReference=Zhang Q HSun Z QHu Wet al.Multi-View Knowledge Graph Embedding for Entity Alignment[C]//Kraus S.Proceedings of the 28th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2019:5429-5435., articleTitle=Multi-View Knowledge Graph Embedding for Entity Alignment, refAbstract=null), Reference(id=1269679041234768586, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2025, volume=24, issue=8, pageStart=17, pageEnd=26, url=null, language=null, rfNumber=14, rfOrder=13, authorNames=杨一帆, 刘忠博, 周玉新, journalName=软件导刊, refType=null, unstructuredReference=杨一帆,刘忠博,周玉新.中医知识图谱构建及应用综述[J].软件导刊202524(8):17-26., articleTitle=中医知识图谱构建及应用综述, refAbstract=null), Reference(id=1269679041318654667, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2015, volume=10, issue=3, pageStart=80, pageEnd=82, url=null, language=null, rfNumber=15, rfOrder=14, authorNames=于彤, 刘静, 贾李蓉, journalName=中国数字医学, refType=null, unstructuredReference=于彤,刘静,贾李蓉,.大型中医药知识图谱构建研究[J].中国数字医学201510(3):80-82., articleTitle=大型中医药知识图谱构建研究, refAbstract=null), Reference(id=1269679041398346444, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=144, pageEnd=147, url=null, language=null, rfNumber=16, rfOrder=15, authorNames=Zhou Y, Qi X L, Huang Y, journalName=null, refType=null, unstructuredReference=Zhou YQi X LHuang Yet al.Research on Construction and Application of TCM Knowledge Graph Based on Ancient Chinese Texts[C]//Barnaghi P,Gottlob G,Katsaros D,et al.Proceedings of the 2019 IEEE International Conference on Bioinformatics and Biomedicine.Piscataway:IEEE,2019:144-147., articleTitle=Research on Construction and Application of TCM Knowledge Graph Based on Ancient Chinese Texts, refAbstract=null), Reference(id=1269679041494815437, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=24, issue=9, pageStart=3613, pageEnd=3621, url=null, language=null, rfNumber=17, rfOrder=16, authorNames=李芊芊, 付兴, 杨凤, journalName=世界科学技术—中医药现代化, refType=null, unstructuredReference=李芊芊,付兴,杨凤,.基于“病脉证并治”诊疗思维的《伤寒论》知识图谱构建与应用[J].世界科学技术—中医药现代化202224(9):3613-3621., articleTitle=基于“病脉证并治”诊疗思维的《伤寒论》知识图谱构建与应用, refAbstract=null), Reference(id=1269679041561924302, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=40, issue=8, pageStart=107, pageEnd=117, url=null, language=null, rfNumber=18, rfOrder=17, authorNames=张卫东, 张晓晓, journalName=情报科学, refType=null, unstructuredReference=张卫东,张晓晓.中医古籍数字资源知识组织与可视化研究——以《金匮要略》为例[J].情报科学202240(8):107-117., articleTitle=中医古籍数字资源知识组织与可视化研究——以《金匮要略》为例, refAbstract=null), Reference(id=1269679041666781903, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=66, issue=22, pageStart=16, pageEnd=27, url=null, language=null, rfNumber=19, rfOrder=18, authorNames=李贺, 祝琳琳, 刘嘉宇, journalName=图书情报工作, refType=null, unstructuredReference=李贺,祝琳琳,刘嘉宇,.基于本体的简帛医药知识组织研究[J].图书情报工作202266(22):16-27., articleTitle=基于本体的简帛医药知识组织研究, refAbstract=null), Reference(id=1269679041738085072, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2025, volume=45, issue=2, pageStart=4, pageEnd=16, url=null, language=null, rfNumber=20, rfOrder=19, authorNames=时倩如, 李贺, 于雯倩, journalName=现代情报, refType=null, unstructuredReference=时倩如,李贺,于雯倩,.中医文本命名实体识别研究综述[J].现代情报202545(2):4-16., articleTitle=中医文本命名实体识别研究综述, refAbstract=null), Reference(id=1269679041830359761, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2025, volume=45, issue=4, pageStart=49, pageEnd=59, url=null, language=null, rfNumber=21, rfOrder=20, authorNames=李明, 罗晓兰, 朱邦贤, journalName=图书馆论坛, refType=null, unstructuredReference=李明,罗晓兰,朱邦贤.中医古籍方剂数据挖掘与知识问答系统构建[J].图书馆论坛202545(4):49-59., articleTitle=中医古籍方剂数据挖掘与知识问答系统构建, refAbstract=null), Reference(id=1269679041897468626, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=1, pageEnd=18, url=null, language=null, rfNumber=22, rfOrder=21, authorNames=许天翼, 石进, journalName=数据分析与知识发现, refType=null, unstructuredReference=许天翼,石进.基于循环-过滤检索增强生成技术的中医知识问答系统研究[J/OL].数据分析与知识发现:1-18[2026-02-02].https://link.cnki.net/urlid/10.1478.G2.20250815.1141.006., articleTitle=基于循环-过滤检索增强生成技术的中医知识问答系统研究, refAbstract=null), Reference(id=1269679042048463571, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2025, volume=45, issue=2, pageStart=26, pageEnd=37, url=null, language=null, rfNumber=23, rfOrder=22, authorNames=沈旺, 于琳, 冯欣, journalName=现代情报, refType=null, unstructuredReference=沈旺,于琳,冯欣,.基于动态知识图谱的中医疫病古籍知识演化研究[J].现代情报202545(2):26-37., articleTitle=基于动态知识图谱的中医疫病古籍知识演化研究, refAbstract=null), Reference(id=1269679042119766740, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2024, volume=65, issue=12, pageStart=1201, pageEnd=1207, url=null, language=null, rfNumber=24, rfOrder=23, authorNames=杨凤, 张艺, 陶晓华, journalName=中医杂志, refType=null, unstructuredReference=杨凤,张艺,陶晓华,.基于知识元理论与临床需求深度融合的中医古籍数智化研究路径和范式[J].中医杂志202465(12):1201-1207., articleTitle=基于知识元理论与临床需求深度融合的中医古籍数智化研究路径和范式, refAbstract=null), Reference(id=1269679042216235733, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2024, volume=30, issue=11, pageStart=172, pageEnd=177, url=null, language=null, rfNumber=25, rfOrder=24, authorNames=齐佳杰, 魏东升, 刘雨婷, journalName=中医药导报, refType=null, unstructuredReference=齐佳杰,魏东升,刘雨婷,.基于“病机链”理念的《脾胃论》湿热成痿肺金受邪论Neo4j知识图谱构建[J].中医药导报202430(11):172-177., articleTitle=基于“病机链”理念的《脾胃论》湿热成痿肺金受邪论Neo4j知识图谱构建, refAbstract=null), Reference(id=1269679042291733206, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2024, volume=47, issue=2, pageStart=178, pageEnd=188, url=null, language=null, rfNumber=26, rfOrder=25, authorNames=张君冬, 杨松桦, 严颖, journalName=情报理论与实践, refType=null, unstructuredReference=张君冬,杨松桦,严颖,.跨医学体系下医疗知识图谱的构建与药物预测研究——以动脉粥样硬化为例[J].情报理论与实践202447(2):178-188., articleTitle=跨医学体系下医疗知识图谱的构建与药物预测研究——以动脉粥样硬化为例, refAbstract=null), Reference(id=1269679042392396503, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=45, issue=6, pageStart=1195, pageEnd=1225, url=null, language=null, rfNumber=27, rfOrder=26, authorNames=张富, 杨琳艳, 李健伟, journalName=计算机学报, refType=null, unstructuredReference=张富,杨琳艳,李健伟,.实体对齐研究综述[J].计算机学报202245(6):1195-1225., articleTitle=实体对齐研究综述, refAbstract=null), Reference(id=1269679042543391448, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2022, volume=468, issue=null, pageStart=97, pageEnd=110, url=null, language=null, rfNumber=28, rfOrder=27, authorNames=Jiang W, Liu Y N, Deng X Y, journalName=Neurocomputing, refType=null, unstructuredReference=Jiang WLiu Y NDeng X Y.Fuzzy Entity Alignment via Knowledge Embedding With Awareness of Uncertainty Measure[J].Neurocomputing2022468:97-110., articleTitle=Fuzzy Entity Alignment via Knowledge Embedding With Awareness of Uncertainty Measure, refAbstract=null), Reference(id=1269679042610500313, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2025, volume=624, issue=null, pageStart=129445, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=28, authorNames=Wang C D, Wang W H, Liang Q Y, journalName=Neurocomputing, refType=null, unstructuredReference=Wang C DWang W HLiang Q Yet al.Local and Global Structure-Aware Contrastive Framework for Entity Alignment[J].Neurocomputing2025624:129445., articleTitle=Local and Global Structure-Aware Contrastive Framework for Entity Alignment, refAbstract=null), Reference(id=1269679042690192090, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=349, pageEnd=357, url=null, language=null, rfNumber=30, rfOrder=29, authorNames=Wang Z C, Lv Q S, Lan X H, journalName=null, refType=null, unstructuredReference=Wang Z CLv Q SLan X Het al.Cross-Lingual Knowledge Graph Alignment via Graph Convolutional Networks[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing.Stroudsburg:Association for Computational Linguistics,2018:349-357., articleTitle=Cross-Lingual Knowledge Graph Alignment via Graph Convolutional Networks, refAbstract=null), Reference(id=1269679042816021211, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2026, volume=194, issue=null, pageStart=108139, pageEnd=null, url=null, language=null, rfNumber=31, rfOrder=30, authorNames=Long M X, Wang J H, Ma J X, journalName=Neural Networks, refType=null, unstructuredReference=Long M XWang J HMa J Xet al.LLM-Augmented Entity Alignment:An Unsupervised and Training-Free Framework[J].Neural Networks2026194:108139., articleTitle=LLM-Augmented Entity Alignment:An Unsupervised and Training-Free Framework, refAbstract=null), Reference(id=1269679042891518684, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=3130, pageEnd=3136, url=null, language=null, rfNumber=32, rfOrder=31, authorNames=Pei S C, Yu L, Hoehndorf R, journalName=null, refType=null, unstructuredReference=Pei S CYu LHoehndorf Ret al.Semi-Supervised Entity Alignment via Knowledge Graph Embedding With Awareness of Degree Difference[C]//Proceedings of the World Wide Web Conference.New York:Association for Computing Machinery,2019:3130-3136., articleTitle=Semi-Supervised Entity Alignment via Knowledge Graph Embedding With Awareness of Degree Difference, refAbstract=null), Reference(id=1269679042962821853, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2023, volume=26, issue=5, pageStart=2265, pageEnd=2290, url=null, language=null, rfNumber=33, rfOrder=32, authorNames=Li Y F, Chen L, Liu C F, journalName=World Wide Web-Internet and Web Information Systems, refType=null, unstructuredReference=Li Y FChen LLiu C Fet al.Generative Adversarial Network for Unsupervised Multi-Lingual Knowledge Graph Entity Alignment[J].World Wide Web-Internet and Web Information Systems202326(5):2265-2290., articleTitle=Generative Adversarial Network for Unsupervised Multi-Lingual Knowledge Graph Entity Alignment, refAbstract=null), Reference(id=1269679043134788318, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=821, pageEnd=832, url=null, language=null, rfNumber=34, rfOrder=33, authorNames=Mao X, Wang W T, Wu Y B, journalName=null, refType=null, unstructuredReference=Mao XWang W TWu Y Bet al.Boosting the Speed of Entity Alignment 10x:Dual Attention Matching Network With Normalized Hard Sample Mining[C]//Proceedings of the 30th World Wide Web Conference.New York:Association for Computing Machinery,2021:821-832., articleTitle=Boosting the Speed of Entity Alignment 10, refAbstract=null), Reference(id=1269679043390640863, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2025, volume=282, issue=null, pageStart=127630, pageEnd=null, url=null, language=null, rfNumber=35, rfOrder=34, authorNames=Meng X Z, Jiao X X, Li J H, journalName=Expert Systems With Applications, refType=null, unstructuredReference=Meng X ZJiao X XLi J Het al.CELLMEA:A Collaboratively Enhanced Large Language Model-Based Entity Alignment for Aircraft Fault Maintenance[J].Expert Systems With Applications2025282:127630., articleTitle=CELLMEA:A Collaboratively Enhanced Large Language Model-Based Entity Alignment for Aircraft Fault Maintenance, refAbstract=null), Reference(id=1269679043474526944, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=4396, pageEnd=4402, url=null, language=null, rfNumber=36, rfOrder=35, authorNames=Sun Z Q, Hu W, Zhang Q H, journalName=null, refType=null, unstructuredReference=Sun Z QHu WZhang Q Het al.Bootstrapping Entity Alignment With Knowledge Graph Embedding[C]//Lang J.Proceedings of the 27th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2018:4396-4402., articleTitle=Bootstrapping Entity Alignment With Knowledge Graph Embedding, refAbstract=null), Reference(id=1269679043533247201, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=796, pageEnd=804, url=null, language=null, rfNumber=37, rfOrder=36, authorNames=Liu B, Lan T C, Hua W, journalName=null, refType=null, unstructuredReference=Liu BLan T CHua Wet al.Dependency-Aware Self-Training for Entity Alignment[C]//Proceedings of the 16th ACM International Conference on Web Search and Data Mining.New York:Association for Computing Machinery,2023:796-804., articleTitle=Dependency-Aware Self-Training for Entity Alignment, refAbstract=null), Reference(id=1269679043612938978, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=5278, pageEnd=5284, url=null, language=null, rfNumber=38, rfOrder=37, authorNames=Wu Y T, Liu X, Feng Y S, journalName=null, refType=null, unstructuredReference=Wu Y TLiu XFeng Y Set al.Relation-Aware Entity Alignment for Heterogeneous Knowledge Graphs[C]//Kraus S.Proceedings of the 28th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2019:5278-5284., articleTitle=Relation-Aware Entity Alignment for Heterogeneous Knowledge Graphs, refAbstract=null), Reference(id=1269679043680047843, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=1452, pageEnd=1461, url=null, language=null, rfNumber=39, rfOrder=38, authorNames=Cao Y X, Liu Z Y, Li C J, journalName=null, refType=null, unstructuredReference=Cao Y XLiu Z YLi C Jet al.Multi-Channel Graph Neural Network for Entity Alignment[C]//Korhonen A,Traum D,Marquez L.Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.Stroudsburg:Association for Computational Linguistics,2019:1452-1461., articleTitle=Multi-Channel Graph Neural Network for Entity Alignment, refAbstract=null), Reference(id=1269679043734573796, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2024, volume=36, issue=6, pageStart=2357, pageEnd=2371, url=null, language=null, rfNumber=40, rfOrder=39, authorNames=Zhang R, Su Y X, Trisedya B D, journalName=IEEE Transactions on Knowledge and Data Engineering, refType=null, unstructuredReference=Zhang RSu Y XTrisedya B Det al.AutoAlign:Fully Automatic and Effective Knowledge Graph Alignment Enabled by Large Language Models[J].IEEE Transactions on Knowledge and Data Engineering202436(6):2357-2371., articleTitle=AutoAlign:Fully Automatic and Effective Knowledge Graph Alignment Enabled by Large Language Models, refAbstract=null), Reference(id=1269679043801682661, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, doi=null, pmid=null, pmcid=null, year=2024, volume=null, issue=null, pageStart=7566, pageEnd=7583, url=null, language=null, rfNumber=41, rfOrder=40, authorNames=Jiang X H, Shen Y H, Shi Z C, journalName=null, refType=null, unstructuredReference=Jiang X HShen Y HShi Z Cet al.Unlocking the Power of Large Language Models for Entity Alignment[C]//Ku L W,Martins A,Srikumar V.Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics.Stroudsburg:Association for Computational Linguistics,2024:7566-7583., articleTitle=Unlocking the Power of Large Language Models for Entity Alignment, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1269679029989839477, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=1, ext=[AuthorCompanyExt(id=1269679030006616694, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1School of Business and Management,Jilin University,Changchun130012,China), AuthorCompanyExt(id=1269679030019199607, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679029989839477, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1吉林大学商学与管理学院,吉林长春130012)]), AuthorCompany(id=1269679030111474296, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, xref=2, ext=[AuthorCompanyExt(id=1269679030124057209, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679030111474296, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Jilin Provincial Library,Changchun130022,China), AuthorCompanyExt(id=1269679030136640122, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, companyId=1269679030111474296, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2吉林省图书馆,吉林长春130022)])], figs=[ArticleFig(id=1269679037648638629, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Fig.1, caption=Entity Alignment Model Integrating Generative Adversarial Learning and Fuzzy Semantic Identification, figureFileSmall=W6o1eA/t5c3r2spsU/MNUQ==, figureFileBig=bkYTxRUx98ANw1fJcXBQVA==, tableContent=null), ArticleFig(id=1269679038017737382, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=图1, caption=融合生成对抗学习与模糊语义辨识的实体对齐模型, figureFileSmall=W6o1eA/t5c3r2spsU/MNUQ==, figureFileBig=bkYTxRUx98ANw1fJcXBQVA==, tableContent=null), ArticleFig(id=1269679038214869671, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Fig.2, caption=Overall Structure and Local Example of Shennong Bencao Jing Knowledge Graph (KG1), figureFileSmall=jnXA8BvVMtxRyv7fSrq4Bg==, figureFileBig=4/SqcwgzQ+h4g1iOhWDYTg==, tableContent=null), ArticleFig(id=1269679038281978536, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=图2, caption=《神农本草经》知识图谱(KG1)的整体结构与局部示例, figureFileSmall=jnXA8BvVMtxRyv7fSrq4Bg==, figureFileBig=4/SqcwgzQ+h4g1iOhWDYTg==, tableContent=null), ArticleFig(id=1269679038382641833, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Fig.3, caption=Overall Structure and Local Example of Tangye Bencao Knowledge Graph (KG2), figureFileSmall=51R2rWG1vbjaJyaWLMEqeg==, figureFileBig=Y+7P0TAO2/+lvEu1GYczyA==, tableContent=null), ArticleFig(id=1269679038521053866, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=图3, caption=《汤液本草》知识图谱(KG2)的整体结构与局部示例, figureFileSmall=51R2rWG1vbjaJyaWLMEqeg==, figureFileBig=Y+7P0TAO2/+lvEu1GYczyA==, tableContent=null), ArticleFig(id=1269679038592357035, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Fig.4, caption=Analysis of Impact of Core Hyperparameters on Model Performance, figureFileSmall=9mHFHrUwxX3udrEGYoVtKg==, figureFileBig=xQKv9YUIN5HGMOG40q0zHQ==, tableContent=null), ArticleFig(id=1269679038672048812, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=图4, caption=核心超参数对模型性能的影响分析, figureFileSmall=9mHFHrUwxX3udrEGYoVtKg==, figureFileBig=xQKv9YUIN5HGMOG40q0zHQ==, tableContent=null), ArticleFig(id=1269679038739157677, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Fig.5, caption=Overall Structure and Local Example of Fused Knowledge Graph SKG-Fused, figureFileSmall=W+iTwUDSMVSHQun4U1MxlQ==, figureFileBig=GDxyo6rGoNI4B7EVmD+jiQ==, tableContent=null), ArticleFig(id=1269679038827238062, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=图5, caption=融合知识图谱SKG-Fused的整体结构与局部示例, figureFileSmall=W+iTwUDSMVSHQun4U1MxlQ==, figureFileBig=GDxyo6rGoNI4B7EVmD+jiQ==, tableContent=null), ArticleFig(id=1269679038898541231, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Fig.6, caption=Example of Knowledge Fusion and Alignment Effects for Baiwei Entity Based on Classic Sources, figureFileSmall=YKQ8WiTRyrZieLtLlPk7Bw==, figureFileBig=PK62y8csubsMpAjOQTenWQ==, tableContent=null), ArticleFig(id=1269679038999204528, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=图6, caption=基于典籍来源构成的白薇实体知识融合与对齐效果示例, figureFileSmall=YKQ8WiTRyrZieLtLlPk7Bw==, figureFileBig=PK62y8csubsMpAjOQTenWQ==, tableContent=null), ArticleFig(id=1269679039066313393, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Tab.1, caption=

Entity Types and Counts in Knowledge Graphs SKG1 and SKG2

, figureFileSmall=null, figureFileBig=null, tableContent=
实体

《神农本草经》

SKG1

《汤液本草》

SKG2

实体对齐

数量

中药529358203
中药别名1 44761\
疾病253378125
症状505746234
功效4951 162218
原文76725\
产地29319\
用法8111\
五味1419\
四性1116\
炮制方法3146\
其他109269\
总计3 7713 910780
), ArticleFig(id=1269679039154393778, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=表1, caption=

知识图谱SKG1和SKG2中的实体类型和数量

, figureFileSmall=null, figureFileBig=null, tableContent=
实体

《神农本草经》

SKG1

《汤液本草》

SKG2

实体对齐

数量

中药529358203
中药别名1 44761\
疾病253378125
症状505746234
功效4951 162218
原文76725\
产地29319\
用法8111\
五味1419\
四性1116\
炮制方法3146\
其他109269\
总计3 7713 910780
), ArticleFig(id=1269679039217308339, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Tab.2, caption=

Model Parameters

, figureFileSmall=null, figureFileBig=null, tableContent=
参数类别参数设定值简述
模型结构嵌入维度768节点由Bert初始化的特征维度
GAT隐藏层维度128图注意力网络的隐藏层大小
注意力头数4GAT中的多头注意力头数
损失函数权重对齐损失间隔1.0FuzzyMarginLoss中的间隔超参数
模糊采样相似度阈值0.8筛选高相似度困难负样本的语义阈值
对抗训练权重0.2整体损失中Ladv_G的平衡权重
对齐与迭代参数伪标签生成阈值0.9迭代过程中筛选高质量对齐样本的相似度门槛
初始最大伪标签数20迭代初期每轮新增的伪标签数量上限
伪标签衰减率0.95每轮迭代中最大伪标签数的动态衰减系数,防止噪声干扰
学习率参数BERT学习率0.000 05BERT的学习速率
GAT及门控学习率0.001GAT编码器与门控融合机制的学习速率
判别器学习率0.000 5对抗网络判别器的学习速率
), ArticleFig(id=1269679039313777332, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=表2, caption=

模型参数

, figureFileSmall=null, figureFileBig=null, tableContent=
参数类别参数设定值简述
模型结构嵌入维度768节点由Bert初始化的特征维度
GAT隐藏层维度128图注意力网络的隐藏层大小
注意力头数4GAT中的多头注意力头数
损失函数权重对齐损失间隔1.0FuzzyMarginLoss中的间隔超参数
模糊采样相似度阈值0.8筛选高相似度困难负样本的语义阈值
对抗训练权重0.2整体损失中Ladv_G的平衡权重
对齐与迭代参数伪标签生成阈值0.9迭代过程中筛选高质量对齐样本的相似度门槛
初始最大伪标签数20迭代初期每轮新增的伪标签数量上限
伪标签衰减率0.95每轮迭代中最大伪标签数的动态衰减系数,防止噪声干扰
学习率参数BERT学习率0.000 05BERT的学习速率
GAT及门控学习率0.001GAT编码器与门控融合机制的学习速率
判别器学习率0.000 5对抗网络判别器的学习速率
), ArticleFig(id=1269679039414440629, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Tab.3, caption=

Comparative Experimental Results of GAFL-Align and Baseline Models

, figureFileSmall=null, figureFileBig=null, tableContent=
模型

Hits@1/

%

Hits@3/

%

Hits@10/

%

MRR/

%

MTransE74.1081.5487.1877.84
GCN-align75.3882.3188.2179.03
RDGCN78.9785.3890.5182.48
MuGNN80.5085.1390.0083.71
BootEA77.6984.1089.4981.75
BERT-INT78.2184.6289.7482.12
LEA82.0585.1391.0384.82
AutoAlign81.2886.4190.7683.98
ChatEA82.8287.4392.0585.35

GAFL-Align

(本模型)

83.5988.9792.5686.23
), ArticleFig(id=1269679039481549494, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=表3, caption=

GAFL-Align与基线模型的对比实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型

Hits@1/

%

Hits@3/

%

Hits@10/

%

MRR/

%

MTransE74.1081.5487.1877.84
GCN-align75.3882.3188.2179.03
RDGCN78.9785.3890.5182.48
MuGNN80.5085.1390.0083.71
BootEA77.6984.1089.4981.75
BERT-INT78.2184.6289.7482.12
LEA82.0585.1391.0384.82
AutoAlign81.2886.4190.7683.98
ChatEA82.8287.4392.0585.35

GAFL-Align

(本模型)

83.5988.9792.5686.23
), ArticleFig(id=1269679039561241271, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Tab.4, caption=

Ablation Study Results of GAFL-Align

, figureFileSmall=null, figureFileBig=null, tableContent=
模型Hits@1/%Hits@3/%Hits@10/%MRR/%
移除Bert语义模块77.1882.3187.1879.94
移除生成对抗网络模块82.3184.8788.7284.02
移除门控机制81.7983.0788.4683.51
移除模糊匹配负采样FuzzyMarginLoss83.0887.6990.7785.46
移除迭代对齐训练策略83.0888.2191.5485.88
GAFL-Align(本模型)83.5988.9792.5686.23
), ArticleFig(id=1269679039653515960, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=表4, caption=

GAFL-Align的消融实验结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型Hits@1/%Hits@3/%Hits@10/%MRR/%
移除Bert语义模块77.1882.3187.1879.94
移除生成对抗网络模块82.3184.8788.7284.02
移除门控机制81.7983.0788.4683.51
移除模糊匹配负采样FuzzyMarginLoss83.0887.6990.7785.46
移除迭代对齐训练策略83.0888.2191.5485.88
GAFL-Align(本模型)83.5988.9792.5686.23
), ArticleFig(id=1269679039741596345, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Tab.5, caption=

Statistics and Analysis of Error Cases

, figureFileSmall=null, figureFileBig=null, tableContent=
错误类型局部字面重叠生僻字干扰知识缺失
数量121724
源实体(SKG1)头眩痛腹中症坚痞结薯蓣
真实对齐(SKG2)头风眩痛腹癥瘕坚积山药
错误预测(SKG2)头肿痛腹中坚痛地胆
错因分析错误项“头肿痛”与源实体共享首尾高频字“头”和“痛”,导致字面相似度虚高。模型未能准确捕捉核心病机词“眩”与“肿”的语义差异。目标实体包含生僻异体字“癥”,BERT将其识别为[UNK],导致关键语义表征缺失。模型受“腹中”一词误导,匹配了错误的候选实体。“薯蓣”与“山药”为古今异名,字面无一字相同。模型仅凭语义表征与结构特征均无法建立有效关联,导致随机性误判。
), ArticleFig(id=1269679039829676730, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=表5, caption=

错误案例统计与分析

, figureFileSmall=null, figureFileBig=null, tableContent=
错误类型局部字面重叠生僻字干扰知识缺失
数量121724
源实体(SKG1)头眩痛腹中症坚痞结薯蓣
真实对齐(SKG2)头风眩痛腹癥瘕坚积山药
错误预测(SKG2)头肿痛腹中坚痛地胆
错因分析错误项“头肿痛”与源实体共享首尾高频字“头”和“痛”,导致字面相似度虚高。模型未能准确捕捉核心病机词“眩”与“肿”的语义差异。目标实体包含生僻异体字“癥”,BERT将其识别为[UNK],导致关键语义表征缺失。模型受“腹中”一词误导,匹配了错误的候选实体。“薯蓣”与“山药”为古今异名,字面无一字相同。模型仅凭语义表征与结构特征均无法建立有效关联,导致随机性误判。
), ArticleFig(id=1269679039896785595, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=EN, label=Tab.6, caption=

Comparison of Network Structural Metrics Between Source and Fused Knowledge Graphs

, figureFileSmall=null, figureFileBig=null, tableContent=
图谱名称实体数量关系数量平均度平均聚类系数图密度LCC占比/%
《神农本草经》SKG13 7717 0113.642 40.006 80.000 92497.74
《汤液本草》SKG23 9106 6163.371 40.082 50.000 85398.68
融合图谱SKG-Fused6 82613 1083.720 90.063 10.000 59398.71
), ArticleFig(id=1269679039968088764, tenantId=1146029695717560320, journalId=1269656373470969926, articleId=1269679023706772060, language=CN, label=表6, caption=

源图谱与融合图谱的网络结构指标统计对比

, figureFileSmall=null, figureFileBig=null, tableContent=
图谱名称实体数量关系数量平均度平均聚类系数图密度LCC占比/%
《神农本草经》SKG13 7717 0113.642 40.006 80.000 92497.74
《汤液本草》SKG23 9106 6163.371 40.082 50.000 85398.68
融合图谱SKG-Fused6 82613 1083.720 90.063 10.000 59398.71
)], attaches=null, journal=Journal(id=1269655921169809475, delFlag=0, nameCn=现代情报, nameEn=Journal of Modern Information, nameHistory1=null, nameHistory2=null, issn=1008-0821, eissn=null, cn=22-1182/G3, coden=null, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=hzZ2kIN9Z2OYfM7GC0mG8g==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Modern Information, journalRemark=null, publicationField=null, createdTime=1780639160676, updatedTime=1780639531872, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Natural Sciences, subjectName=null, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=hzZ2kIN9Z2OYfM7GC0mG8g==, picEn=za/Q6fknju6Wnm9GGKKQHw==, jcr=null, cjcr=null, exts=[JournalExt(id=1269657478171099337, language=CN, name=现代情报, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1780639531891, updatedTime=1780639531891, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://xdqbauthor.manuscriptcloud.com/, submissionEditorUrl=https://xdqbeditor.manuscriptcloud.com/, submissionReviewUrl=https://xdqbauthor.manuscriptcloud.com/, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1269657478221430986, language=EN, name=Journal of Modern Information, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1780639531903, updatedTime=1780639531903, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://xdqbauthor.manuscriptcloud.com/, submissionEditorUrl=https://xdqbeditor.manuscriptcloud.com/, submissionReviewUrl=https://xdqbauthor.manuscriptcloud.com/, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1269656373470969926, websiteList=[Website(id=1269658134734860575, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1269656373470969926, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/xdqb/CN, language=CN, createTime=1780639688429, createBy=18614031015, updateTime=1780639910823, updateBy=18614031015, name=现代情报-中文, tplId=1146099689490845704, title=现代情报, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1269659212566131092, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=articleTextType, value=kx, createTime=1780639945403, updateTime=1780639945403, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212540965265, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=banner, value=null, createTime=1780639945397, updateTime=1780639945397, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212587102615, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=grayFlag, value=0, createTime=1780639945408, updateTime=1780639945408, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212532576656, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=logo, value=https://castjournals.cast.org.cn/joweb/xdqb/CN/file/pic?fileId=hAchGHMHdO6KzdfJ3HY9JQ==, createTime=1780639945395, updateTime=1780639945395, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212599685529, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=minRunFlag, value=0, createTime=1780639945411, updateTime=1780639945411, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212557742483, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/xdqb/CN/file/pic, createTime=1780639945401, updateTime=1780639945401, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212595491224, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=silenceFlag, value=0, createTime=1780639945410, updateTime=1780639945410, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212549353874, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1780639945399, updateTime=1780639945399, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212574519701, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=themeColor, value=null, createTime=1780639945405, updateTime=1780639945405, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659212578714006, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134734860575, code=themeStyle, value=null, createTime=1780639945406, updateTime=1780639945406, creator=18614031015, updator=18614031015)]), Website(id=1269658134877466914, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1269656373470969926, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/xdqb/EN, language=EN, createTime=1780639688462, createBy=18614031015, updateTime=1780639906236, updateBy=18614031015, name=现代情报-英文, tplId=1146101810881728533, title=Journal of Modern Information, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1269659152486920583, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=articleTextType, value=kx, createTime=1780639931079, updateTime=1780639931079, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152465949060, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=banner, value=null, createTime=1780639931074, updateTime=1780639931074, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152507892106, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=grayFlag, value=0, createTime=1780639931084, updateTime=1780639931084, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152453366147, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=logo, value=https://castjournals.cast.org.cn/joweb/xdqb/EN/file/pic?fileId=hAchGHMHdO6KzdfJ3HY9JQ==, createTime=1780639931071, updateTime=1780639931071, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152524669324, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=minRunFlag, value=0, createTime=1780639931088, updateTime=1780639931088, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152478531974, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/xdqb/EN/file/pic, createTime=1780639931077, updateTime=1780639931077, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152516280715, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=silenceFlag, value=0, createTime=1780639931087, updateTime=1780639931087, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152474337669, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1780639931076, updateTime=1780639931076, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152495309192, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=themeColor, value=null, createTime=1780639931081, updateTime=1780639931081, creator=18614031015, updator=18614031015), WebsiteProps(id=1269659152503697801, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1269658134877466914, code=themeStyle, value=null, createTime=1780639931083, updateTime=1780639931083, creator=18614031015, updator=18614031015)])], journalTitle=现代情报, weixinUrl=null, journalUrl=https://www.xdqb.net/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Modern Information, journalPhotoCn=hzZ2kIN9Z2OYfM7GC0mG8g==, journalPhotoEn=za/Q6fknju6Wnm9GGKKQHw==, journalFirstLetter=J, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/xdqb/CN/10.3969/j.issn.1008-0821.2026.03.003, detailUrlEn=https://castjournals.cast.org.cn/joweb/xdqb/EN/10.3969/j.issn.1008-0821.2026.03.003, pdfUrlCn=https://castjournals.cast.org.cn/joweb/xdqb/CN/PDF/10.3969/j.issn.1008-0821.2026.03.003, pdfUrlEn=https://castjournals.cast.org.cn/joweb/xdqb/EN/PDF/10.3969/j.issn.1008-0821.2026.03.003, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
面向知识融合的本草典籍知识图谱实体对齐研究
收藏切换
PDF下载
李贺 1, * , 邵文诗 1 , 刘嘉宇 1 , 张津源 1 , 沈旺 1 , 王桂敏 2
现代情报 | 数据智能与知识服务 2026,46(3): 30-43
收起
收藏切换
现代情报 | 数据智能与知识服务 2026, 46(3): 30-43
面向知识融合的本草典籍知识图谱实体对齐研究
全屏
李贺1, *, 邵文诗1, 刘嘉宇1, 张津源1, 沈旺1, 王桂敏2
作者信息
  • 1吉林大学商学与管理学院,吉林长春130012
  • 2吉林省图书馆,吉林长春130022
  • 邵文诗(1999-),女,硕士研究生,研究方向:中医古籍数字人文

    刘嘉宇(1995-),男,助理研究员,博士,研究方向:数据挖掘

    张津源(2001-),女,博士研究生,研究方向:中医古籍数字人文

    沈旺(1983-),女,教授,博士,博士生导师,研究方向:用户信息行为

    王桂敏(1976-),女,研究馆员,硕士,研究方向:科技信息服务。

通讯作者:

李贺(1964-),女,教授,博士,博士生导师,研究方向:信息行为分析、知识管理。
Research on Entity Alignment in Chinese Materia Medica Knowledge Graphs for Knowledge Fusion
He Li1, *, Wenshi Shao1, Jiayu Liu1, Jinyuan Zhang1, Wang Shen1, Guimin Wang2
Affiliations
  • 1School of Business and Management,Jilin University,Changchun130012,China
  • 2Jilin Provincial Library,Changchun130022,China
出版时间: 2026-03-01 doi: 10.3969/j.issn.1008-0821.2026.03.003
文章导航
收藏切换
目的/意义

针对本草典籍知识图谱实体对齐任务中图谱异构、术语易混淆及高质量标注稀缺等挑战,提出融合生成对抗网络与模糊语义辨识的实体对齐模型GAFL-Align,旨在实现多源知识自动化融合。

方法/过程

该模型通过BERT与图注意力网络融合实体语义与拓扑结构,利用生成对抗网络进行领域自适应以消除异构引发的特征分布差异,采用模糊边界负采样策略强化对易混淆术语的细粒度辨识,并结合迭代自训练机制利用高置信度结果扩充样本,有效降低对人工标注的依赖。

结果/结论

实验表明,该模型在自建数据集上的核心指标均优于基线方法。在此基础上构建的多源融合图谱实现了典籍间知识的互补与增值,为本草典籍知识自动化融合提供了有力的技术支撑。

知识融合  /  实体对齐  /  本草典籍  /  知识图谱  /  深度学习
Purpose/Significance

The digital transformation of Chinese Materia Medica(CMM) classics is critical for bridging ancient pharmacological wisdom with modern drug discovery. However, existing Knowledge Graphs(KGs) for CMM are often constructed in isolation, resulting in fragmented information silos that hinder global data interoperability. While Entity Alignment(EA) has become a focal point in the international Semantic Web community, specific research targeting the alignment of ancient CMM literature remains a significant gap. Moreover, current state-of-the-art models—primarily designed for modern, high-resource languages—struggle to address the unique challenges of ancient Chinese texts. These challenges include severe structural heterogeneity caused by disparate historical writing styles, high terminological ambiguity where distinct medical concepts share similar characters, and a critical scarcity of high-quality annotated data⁃sets. This study aims to fill this gap by proposing a domain-specific deep learning framework designed to automate the fusion of multi-source historical medical knowledge.

Method/Process

To overcome these barriers, this paper proposed the Generative Adversarial Fuzzy-boundary Learning(GAFL-Align) model. The study utilized two representative classics spanning different historical eras: Shennong Bencao Jing and Tangye Bencao. After data cleaning, the datasets comprised 3 771 and 3 910 normalized entities, respectively, focusing on core categories such as herbs, symptoms, and diseases. The technical architecture integrated BERT for deep semantic encoding with Graph Attention Networks(GAT) to capture topological structures. To handle distribution shifts across heterogeneous texts, the model employed a Generative Adversarial Network(GAN) for domain adaptation, mapping entities into a unified feature space. Furthermore, a novel fuzzy boundary negative sampling strategy was developed to distinguish “hard negatives”—terms with high lexical similarity but distinct medical meanings. To address data scarcity, an iterative self-training mechanism with confidence-aware filtering was implemented to augment the training set from a limited number of expert-annotated seed pairs.

Result/Conclusion

Experimental results indicated that GAFL-Align achieved a Hits@1 score of 83.59%, significantly outperforming nine baselines, including translation-based models, GNN variants, and Large Language Models(LLMs)-augmented approaches like ChatEA. The model successfully constructed a fused KG containing 6 826 entities, effectively merging heterogeneous data while preserving unique source-specific attributes. These findings demonstrate that combining adversarial domain adaptation with fine-grained semantic differentiation offers a superior solution for low-resource historical knowledge fusion compared to generic LLMs. Ultimately, this research provides a robust technical foundation for the intelligent organization of CMM heritage, offering significant implications for digital humanities and the global standardization of traditional medicine data.

knowledge fusion  /  entity alignment  /  Chinese Materia Medica classics  /  knowledge graph  /  deep learning
李贺, 邵文诗, 刘嘉宇, 张津源, 沈旺, 王桂敏. 面向知识融合的本草典籍知识图谱实体对齐研究. 现代情报, 2026 , 46 (3) : 30 -43 . DOI: 10.3969/j.issn.1008-0821.2026.03.003
He Li, Wenshi Shao, Jiayu Liu, Jinyuan Zhang, Wang Shen, Guimin Wang. Research on Entity Alignment in Chinese Materia Medica Knowledge Graphs for Knowledge Fusion[J]. Journal of Modern Information, 2026 , 46 (3) : 30 -43 . DOI: 10.3969/j.issn.1008-0821.2026.03.003
中医本草典籍承载着数千年的传统医学智慧,其丰富的药物知识体系既为中医临床实践提供理论指导,也为现代药物研发与药理研究提供重要参考1。为有效利用这些宝贵资源,将海量、异构的典籍文献转化为结构化的知识图谱已成为一项关键研究路径2。然而,现有的本草典籍知识图谱往往局限于单一典籍或特定的研究视角来构建3。这种各自独立的构建模式虽具有局部价值,却切断了典籍间的内在联系,导致中医知识体系碎片化,不仅阻碍跨典籍知识的融通,也使基于全局视角的智能推理与知识发现难以实现4。因此,整合分散的知识资源,构建更为全面、一致的本草典籍知识图谱,已成为该领域的迫切需求5。实体对齐(Entity Alignment)正是应对此挑战的关键技术,主要通过识别并链接不同图谱中指代同一现实世界对象的实体,为实现多源知识的有效融合与互补提供了核心方法6
尽管实体对齐技术在通用领域已取得显著进展,但将其直接应用于中医本草典籍时,仍面临诸多领域特有的挑战。首先,图谱结构异构问题限制了模型的鲁棒性。不同典籍在著述风格、术语体系乃至图谱拓扑结构上存在显著差异7,而现有模型往往忽视这种跨图谱分布偏差8-9,导致对齐精度受限。其次,术语易混淆是该领域的另一大难题10。本草典籍中充斥着大量形近义远的困难样本,使得现有依赖随机负采样策略的模型容易产生误判11-13,难以精准捕捉细微的语义差别。最后,高质量标注数据的稀缺成为制约技术应用的主要瓶颈4。由于本草领域的专业门槛导致标注成本较高,依赖大规模监督信号的传统深度学习方法在低资源环境下往往难以奏效。
针对上述挑战,本文提出一种融合生成对抗网络与模糊语义辨识的实体对齐模型(Generative Adversarial Fuzzy-boundary Learning,GAFL-Align),旨在通过跨域特征自适应机制、模糊边界负采样策略(Fuzzy Negative Sampling)以及迭代自训练方法(Bootstrapping),攻克本草典籍实体对齐中图谱异构、术语混淆及标注数据稀缺的难题。在此基础上,本研究构建了一个融合知识图谱,有效实现了不同典籍间异构知识的互补与增值,为本草典籍知识的系统性整合提供了可行的技术路径。
近年来,面向本草典籍的知识图谱研究主要围绕知识图谱的构建与应用两大方向展开14。在构建方面,早期研究侧重于领域本体的构建15-16以及依赖专家经验的手工知识抽取17-18。随着对构建规模与效率要求的不断提升,研究重心逐渐转向自动化知识抽取技术。历经了从以条件随机场为代表的传统机器学习方法的探索19,到以BERT-BiLSTM-CRF为代表的深度学习模型的引入20,并且随着大语言模型(Large Language Models,LLMs)的兴起,以此为驱动的全自动化构建方式亦开始浮现21。而在应用方面,基于已构建的知识图谱,研究者在智能问答22、可视化分析23和辅助决策24等方面开展了诸多有益尝试,拓展了本草典籍知识在实际场景中的使用价值。
然而,当前本草典籍知识图谱研究或局限于单一典籍19,知识覆盖严重不足;或在整合多部典籍时仅停留在表层数据的汇集21-23,同一实体常以不同名称在图谱中冗余且割裂地存在,破坏了知识图谱的语义一致性。知识融合特别是实体对齐在该领域的研究相对滞后,少数对齐尝试主要停留在人工对齐25或基于表层规则与字符串匹配26的阶段,难以应对本草典籍中复杂的语义现象,迫切需要引入能够理解深层语义的智能化自动实体对齐模型。
实体对齐旨在识别并链接不同知识图谱中的等价实体,是实现知识图谱融合的关键技术27。早期的研究主要基于以TransE及其变体为代表的翻译模型28,如MTransE11等。随着图神经网络的兴起,基于图结构的对齐方法成为主流29。GCN-Align模型30首次利用GCN通过聚合邻域信息来生成实体的结构化嵌入。随后,RREA模型8引入关系感知机制,显著增强了实体在不同关系下的语义表征能力。为弥补单一结构信息的不足,语义增强成为另一重要趋势。例如,BERT-INT模型9通过微调BERT捕获细粒度文本语义。近期,利用大语言模型强大的泛化与推理能力来提升对齐效果,已成为这一领域的重要探索方向。例如,LEA模型31通过LLM增强实体的文本化语义描述,将对齐转化为检索任务。
在基础架构之外,针对实体对齐中特定难题的优化策略也相继被提出。首先,在嵌入特征优化方面,对抗学习机制被引入以试图缓解分布异构问题。其中,SEA模型32主要致力于减少由实体度分布差异引发的嵌入偏差,而GAEA模型33则通过模拟伪结构分布进行对抗训练,旨在改善无监督场景下的噪声干扰问题。其次,针对负采样质量不高的问题,研究者提出了从困难样本挖掘34到自适应负样本混合35等策略,旨在提升模型对困难样本的辨识能力。最后,为缓解数据稀疏问题,迭代自训练机制被广泛采用,BootEA模型36率先确立了利用高置信度预测结果扩充训练集的自训练方法,而后续研究如STEA模型37则改进了数据扩充策略,通过引入实体依赖关系来校验预测结果的结构兼容性,进而生成高质量的依赖感知伪标签以持续丰富训练样本。
针对本草典籍领域的深层语义对齐研究尚处于起步阶段,直接迁移现有模型仍面临诸多局限。具体而言,现有对抗模型多聚焦局部图结构的去噪,未能从典籍源异构的宏观视角,有效消除因著述风格与时代变迁引发的整体特征分布的差异,致使模型难以建立跨典籍实体间的准确映射关系。同时,针对形近义远术语的细粒度辨识机制依然缺失,通用的随机负采样策略往往难以挖掘处于语义模糊边界的困难样本,无法满足本草知识对齐的精确性要求。此外,本草典籍实体对齐受制于繁重的人工校验与复杂的规则构建,尚未引入迭代自训练等半监督机制以挖掘未标注数据的价值,限制了模型在低资源场景下的泛化能力。
本文设计并构建了一个融合生成对抗学习与模糊语义辨识的实体对齐模型GAFL-Align,如图1所示。该模型首先利用BERT(Bidirectional Encoder Representations From Transformers)预训练语言模型捕捉本草典籍术语的深层语义信息,生成富含先验知识的节点初始特征。随后将图注意力网络(Graph Attention Networks,GAT)作为结构编码器与GAN生成器,先深度聚合实体的语义内容与图谱的拓扑结构生成嵌入表示,继而通过与判别器的动态对抗博弈,消除由不同典籍著述风格带来的特征分布差异,将实体映射到统一的特征空间中。与此同时,针对本草典籍中大量形近义远的易混淆术语,本研究设计了模糊边界负采样策略,迫使模型关注语义空间中距离极近但非同义的困难样本,以提升对细粒度语义差异的辨识能力。最后,结合迭代自训练策略,利用模型的高置信度预测结果自动扩充训练样本,从而有效缓解对大规模人工标注数据的依赖。
本文将本草典籍知识图谱定义为G=(E,R,T),其中,E代表实体集,R代表关系集,TE × R × E表示三元组集合,三元组(h,r,t)T表示头实体h通过关系r连接到尾实体t。给定两个来源不同且异构的本草知识图谱KG1=(E1,R1,T1)KG2=(E2,R2,T2),实体对齐的任务是发现两个图谱中指代同一现实对象的等价实体对集合A。该集合定义为式(1):
A={(e1,e2)|e1E1,e2E2,e1e2}
其中,表示等价关系。在本研究的半监督设置下,假设已有一组少量的预对齐实体对作为训练种子集StrainA,模型的目标是利用KG1KG2Strain自动识别出剩余未知的对齐实体对。
为精确捕捉本草典籍术语名称中蕴含的深层语义信息,本模型采用BERT预训练语言模型对所有图节点进行初始特征向量化。对于知识图谱中的任一节点v,其文本名称textv,输入至预训练的bert-base-chinese模型,并提取与分类任务相关的[CLS]标记所对应的输出向量,作为该节点的初始特征向量Xv,该过程形式化表示为式(2):
Xv=BERT(textv)
其中,Xvd,在本研究中嵌入维度d设定为768。通过此步骤,模型将本草典籍实体映射为连续的高维稠密向量,不仅解决了特定术语的语义表示问题,更为后续图神经网络融合提供了高质量的语义特征。
仅依赖BERT提取的文本语义不足以完全表示实体全部信息,因为在本草典籍知识图谱中,结构上下文蕴含着关键的判别信息。考虑到不同邻居节点对实体对齐的贡献度存在差异,本研究采用GAT作为结构编码器。GAT既起到深度融合语义与图谱结构的功能,又作为生成对抗网络中的生成器。
作为编码器,该模块旨在实现文本语义与拓扑结构的深度融合。其接收由BERT初始化的语义特征矩阵X与图谱邻接矩阵A作为输入。GAT利用自注意力机制,根据节点间语义特征的关联强度动态分配结构聚合的权重。对于任意节点i及其邻居j,其注意力系数αij计算如式(3)所示:
αij=exp (LeakyReLU(aT[WhiWhj]))k𝒩iexp (LeakyReLU(aT[WhiWhk]))
其中,W为可学习的权重矩阵,a为注意力向量,表示向量拼接操作。
作为生成器,模型利用上述注意力系数对邻域特征进行加权聚合,直接构建用于对抗博弈的全局特征矩阵Z。对于矩阵中的任意行向量ziZ,其计算公式定义为式(4):
zi=σj𝒩iαijWhj, ziZ
尽管前述GAT模块能够有效融合局部结构信息,但由于不同典籍成书年代悬殊,其著述风格、用词习惯及潜在的图拓扑特征存在差异。这种异构性导致生成的实体嵌入ZKG1ZKG2在特征空间中服从不同的概率分布,若直接在此基础上计算相似度,这种分布偏差将掩盖实体间真实的语义关联,导致对齐精度受限。为此,本节引入基于对抗性领域自适应(Adversarial Domain Adaptation)机制,旨在迫使模型学习到一个与具体典籍来源无关的统一特征空间。为了消除KG1KG2之间的整体分布差异,本模型引入了一个二分类判别器D。该判别器由多层感知机构成,其目标是尽可能精准地分辨输入的实体嵌入Zi是来自源图谱还是目标图谱。训练过程遵循最小化—最大化博弈策略,判别器通过最小化二元交叉熵损失来进行优化,其损失函数adv_D定义为式(5):
adv_D=-𝔼Zi{ZKG1  ZKG2}cilog D(Zi) + (1 - ci)log (1 - D(Zi))
其中,ci为实体的真实来源标签。与之相对,生成器的目标则是生成具有混淆性的嵌入以误导判别器,其对抗损失adv_G定义为式(6):
adv_G=-𝔼Zi{ZKG1  ZKG2}(1 - ci)log D(Zi) + cilog (1 - D(Zi))
通过上述adv_Dadv_G的交替优化,模型最终达到纳什均衡状态。此时,生成的实体嵌入不仅保留了语义与结构特征,还消除了由典籍来源引起的分布差异,从而将异构的KG1KG2映射到了一个统一的向量空间中。
经过前述对抗训练优化后,生成器输出的嵌入Z虽然消除了图谱间的分布差异,具备良好的跨域不变性,但这种为了迎合判别器而进行的对抗性调整,可能会在一定程度上导致原始实体特征丢失了部分对于精准对齐至关重要的细粒度语义信息。
为解决消除图谱异构与保留语义信息之间的潜在冲突,本模型引入了一个自适应门控融合机制,动态权衡原始高保真语义特征X与对抗适应后的结构特征Z。门控系数g通过式(7)学习:
g=σ(Wg[XZ] + bg)
其中,表示向量拼接操作,Wgbg为可学习的权重与偏置,σ为Sigmoid激活函数。随后,利用该门控系数对两种特征进行加权融合,生成最终用于对齐任务的实体嵌入Zalign,如式(8)所示:
Zalign=gX + (1 - g)Z
其中,表示元素级乘法。通过这一机制,模型能够智能地对特征进行选择,对于语义歧义较大但结构清晰的实体,模型会自动赋予Z更高的权重,而对于那些语义独特但结构噪声较大的实体,则更多地保留X中的原始信息,最终得到的Zalign兼具了语义丰富性与分布一致性。
尽管前述生成对抗网络有效解决了宏观图谱异构,但其忽略了微观细粒度语义差异,且传统随机负采样因引入大量简单负样本,难以区分特征空间中极近的混淆项。为此,本研究提出模糊边界负采样策略,旨在挖掘语义模糊地带的困难负样本,将优化重心转向局部精细化边界,迫使模型学习细粒度辨识能力。
具体而言,模型针对任意锚点实体ei构建模糊负样本候选集。该过程利用BERT初始语义向量计算锚点与图谱中其他实体的余弦相似度,并引入语义阈值来划定模糊边界。在此筛选过程中,算法排除了真实的对齐目标,以防止将正样本误标记为负例而干扰模型优化。即只有相似度高于该阈值且并非正确对齐的实体,方可被纳入模糊负样本候选集Sneg(i),其公式如式(9)所示:
Sneg(i)={ekE|sim(BERT(ei),BERT(ek))>τ  ekei  (ei,ek)A}
其中,sim()表示余弦相似度,A为已知的对齐集合。在此基础上,为了最大化梯度的判别效力,模型在训练过程中实施Top-K困难样本挖掘策略。不同于随机策略,该算法对Sneg(i)中的实体按相似度进行降序排列,并截取前K个实体作为困难负样本。这种机制确保了模型始终聚焦那些外观最相似但实质不同的干扰项,迫使模型在统一的特征空间中必须捕捉更细微的语义差异才能完成区分,从而缓解了复杂本草术语的细粒度辨识难题。
为了实现模型中不同组件的协同优化,本研究设计了基于分层梯度的交替迭代训练策略。基于筛选出的Top-K困难负样本,本研究构建了基于间隔的排序损失函数(Margin-based Ranking Loss),旨在拉近正样本对在特征空间中的距离,同时推远处于模糊边界的困难负样本。其形式化定义为式(10):
align=(i,j)Strainmax 0,m + d(Zalign_i,Zalign_j) - d(Zalign_i,Zalign_k)
其中,(i,j)为正样本对,k为候选集Sneg(i)中选取的困难负样本,d()L2欧氏距离度量,m为预设的对齐间隔参数,其核心作用是强制要求负样本对的距离至少要比正样本对的距离大m,从而确保模型在特征空间中学习到足够显著的决策边界。
在参数更新阶段,判别器与生成器采取交替优化的方式。判别器D仅依据二元交叉熵对抗损失adv_D进行独立更新,在此过程中生成器、门控及BERT参数保持固定。与之相对,生成器GAT参数更新由总损失GAT驱动,如式(11)所示:
LGAT=align + λadvadv_G
其中,λadv是用于平衡语义保真度与结构自适应性的超参数。为了防止对抗训练破坏关键的语义辨识能力,本研究在联合更新阶段实施了梯度隔离策略,BERT语义编码模块与自适应门控网络的参数更新仅由align驱动,截断了来自对抗损失adv_G的梯度回传。这种分层梯度的交替优化策略,有效避免了领域自适应过程可能造成的关键语义信息丢失,确保模型在消除典籍异构性的同时,依然能够精准捕捉本草术语的细粒度语义特征。
为缓解标注依赖并挖掘潜在的对齐信息,本研究引入了基于置信度感知的迭代自训练策略,通过训练、预测、扩充的循环机制,利用模型生成的高置信度伪标签逐步扩充训练集。在第t轮优化后,计算未标注实体对的嵌入相似度作为置信度指标。为平衡扩充训练规模与防止噪声引入,本研究设计了动态衰减策略。鉴于迭代后期剩余样本多为语义模糊的困难样本,若维持固定扩充量易引入噪声,故设定初始上限Nmax与衰减系数η。每轮新增伪标签除了需要满足相似度阈值外,数量还受到上限Nmax × ηt的约束。该策略构建了从易到难的课程学习过程,前期快速积累,后期保守扩充以降低噪声污染。筛选出的伪标签集Spseudo(t)将被合并入下一轮训练,如式(12)所示:
Strain(t + 1)=Strain(t)Spseudo(t)
迭代持续至达到预设轮数或无新伪标签生成。该机制实现了从少量已知数据向未知复杂数据的自动知识迁移,提升了在低资源场景下的鲁棒性。
本研究选取《神农本草经》与《汤液本草》作为数据来源,二者成书年代跨度极大,具有显著的特征差异与互补价值,是验证跨典籍知识融合的理想对象。在研究初期,首先定义了统一的中医药本体规范,明确了实体与关系的类型定义。随后,由具备中医药背景的专业人员借助Label Studio平台对原始文献进行人工标注,构建了两个初始知识图谱。其中,源自《神农本草经》的KG1包含4 173个实体,源自《汤液本草》的KG2包含4 582个实体,其整体网络结构及局部拓扑特征如图2图3所示。
为保留真实挑战同时确保实验数据的质量,本研究仅实施了基础的术语清洗工作。结合权威中医词典与盘古分词工具,对原始数据进行了去噪与初步规范化,构建了两个标准化的知识图谱SKG1与SKG2,其实体类型的详细分布如表1所示。
鉴于“原文”实体缺乏重叠、“中药别名”数量差异悬殊,且“五味”与“四性”语义单一易导致模型指标虚高,因此本研究选取“中药”“症状”“疾病”和“功效”四类核心实体构建数据集。由两名中医药专家手工标注构建了780对高质量对齐种子对的标准数据集,其Cohen’s Kappa一致性系数达到了0.86,表明标注质量良好且高度一致。尽管相比通用领域百万级的对齐数据集,该数据集规模较小,但这客观反映了本草典籍领域普遍面临的低资源与高标注成本的双重挑战。这也正是本研究致力于验证模型在小样本或低资源场景下有效性的重要前提。本研究采取50%的划分策略,将种子数据随机均分为两组,其中390对作为初始训练集Strain用于引导迭代自训练,剩余390对作为测试集用于客观评估模型的泛化性能。
实验基于PyTorch框架,在NVIDIA GeForce RTX 3090 GPU上进行。模型训练采用Adam优化器,批次大小为128,最大迭代轮数设为100,并严格实施早停机制(Early Stopping)以防止模型过拟合。核心超参数均经网格搜索寻优确定,具体参数配置如表2所示。
本研究采用实体对齐领域的通用指标Hits@k (k=1,3,10)和平均倒数排名(Mean Reciprocal Rank,MRR)评估模型性能。其中,Hits@1直观反映对齐的准确率,Hits@3与Hits@10考察候选列表的覆盖能力,MRR则衡量整体排序质量。
为评估GAFL-Align的有效性,本研究选取了9个涵盖实体对齐技术演进历程的基线模型,旨在考察其在结构异构、语义复杂及低资源场景下的综合竞争力。其中,MTransE11作为平移距离模型的基准,GCN-align30、RDGCN38与MuGNN39代表了图神经网络的不同变体,引入此类模型旨在验证深层聚合多视角拓扑结构信息对于克服典籍结构异构的重要性。在语义与半监督方面,选取BERT-INT9以评估脱离图结构约束下纯文本语义对齐的独立贡献,选取BootEA36作为迭代自训练的经典代表,用于衡量在缺乏噪声控制机制时半监督策略的实际收益。此外,本研究还引入了LEA31、AutoAlign40与ChatEA41这三类大语言模型增强的前沿方法,分别代表了基于语义检索、零样本全自动对齐以及基于思维链推理的方法。
本研究将GAFL-Align模型与上述基线模型在自建的本草典籍数据集上进行了对比实验,实验结果如表3所示。
从整体实验结果来看,GAFL-Align模型在所有核心指标上均取得了最优性能,显著超越了所有基线方法。通过横向对比基于结构的方法可以发现,以MTransE为代表的翻译模型表现相对最弱,这主要归因于简单的线性变换假设难以有效处理本草典籍中复杂的非线性映射关系。相比之下,GCN-align、RDGCN和MuGNN等图神经网络变体的性能呈现出阶梯式上升的趋势。其中,MuGNN利用多通道机制融合了多视角信息,将Hits@1指标提升至80.50%。这一趋势证明了在本草典籍实体对齐任务中,深度挖掘和聚合邻域拓扑结构信息对于弥补单一典籍的描述缺失具有决定性作用。而BERT-INT模型虽然利用了强大的预训练语义特征,但其78.21%的Hits@1指标仍低于部分GNN模型。这说明脱离了图谱结构约束的纯语义匹配,极易受困于古籍中名异实同或名同实异的语言模糊性陷阱。同时,经典的迭代自训练模型BootEA表现一般,其Hits@1指标为77.69%,这表明在缺乏有效噪声控制机制的情况下,直接引入伪标签容易导致错误累积。这一对比反证了GAFL-Align不仅通过BERT与GAT实现了语义与结构的深层互补,还通过置信度感知的动态迭代策略,规避了传统半监督学习中固有的错误扩散问题。
在大语言模型相关方法的对比中,ChatEA虽凭借强大的语义推理能力,在基线模型中表现最佳,优于LEA和AutoAlign。但GAFL-Align依然保持了约0.77%的优势。这种专用小模型超越通用大模型的现象,主要归功于本模型针对领域特性的深度定制。LLM方法虽然通识语义能力极强,但往往缺乏对本草典籍垂直领域知识图谱异构结构的感知能力,容易忽略隐式的拓扑关联。而GAFL-Align通过生成对抗网络主动消除了跨典籍的特征分布差异,并利用模糊边界负采样策略强迫模型在细粒度上区分形近义远的困难样本。这种针对性的领域自适应设计,使其在处理本草典籍特有的异构与混淆挑战时,比通用大语言模型具有更高的精确度。
为了验证GAFL-Align模型中各个核心组件的有效性,本文设计了消融实验,实验结果如表4所示。
表4的实验数据可以观察到,GAFL-Align模型在四项评估指标上均优于移除了任一单一组件的变体模型。当移除BERT语义模块后,模型性能下降最为显著。这表明,由BERT提供的深层语义信息是模型理解本草典籍实体内涵的基石,缺乏高质量的语义输入,后续的结构和对抗学习将无从谈起。移除门控机制导致的性能降幅位居第二,证明了简单的特征拼接无法替代动态权衡策略,门控机制保障了模型在消除分布差异的同时保留关键语义细节。而移除生成对抗网络模块导致的模型性能下降,则印证了对抗学习在宏观层面上消除典籍著述风格差异、缓解图谱结构异构方面的核心作用。此外,当用常规随机负采样替换模糊边界负采样损失后,模型性能也出现了下降。这证实了针对本草领域形近义远术语设计的模糊采样策略,对于提升模型在处理易混淆实体时的精细分辨能力至关重要。最后,移除迭代对齐训练策略同样导致了性能的降低,验证了半监督学习思想在低资源场景下的应用价值。
本研究选取模糊负采样语义阈值、伪标签置信度阈值以及对抗生成网络权重3个核心参数进行敏感性分析,以验证GAFL-Align模型在不同超参数配置下的鲁棒性,实验结果如图4所示。针对模糊负采样语义阈值,模型性能随该值提升呈现先升后降趋势,并在0.8处达到峰值。这表明过低的阈值难以筛选出具有梯度的困难样本,过高的阈值则导致有效负样本匮乏,限制了模糊边界的优化效果。在伪标签置信度阈值方面,模型在参数设为0.9时表现最优。较低的阈值会引入噪声伪标签导致误差累积,过高的阈值则大幅缩减了训练样本的扩充规模,削弱了自训练收益。此外,对抗生成网络权重在0.2时取得了最佳平衡。权重过小无法有效消除典籍间的分布异构,权重过大则会使原始语义丢失,破坏实体原始特征结构。综上所述,适度的参数设定能有效平衡样本质量与模型约束,验证了模型在特定参数区间内的稳定性。
为深入探究GAFL-Align模型的误差边界与性能瓶颈,本节对测试集中典型Top-1预测错误案例进行了分类统计与定性分析,如表5所示。
尽管消融实验已证实模糊边界负采样策略降低了本草领域形近义远术语的整体误判率,但在处理字面重叠度极高的长尾样本时,模型性能仍会受限。首先,当候选实体与源实体出现多个相同字符时,BERT的注意力机制倾向于聚焦这些共现特征,从而在一定程度上掩盖了核心差异字的语义。但此类错误的出现,恰恰印证了模型已高效剔除了绝大多数常规的中低难度混淆干扰。其次是生僻字干扰造成的识别盲区,由于通用BERT模型自身的词表存在一定的局限性,本草典籍中部分异体字常被编码为未知标记,导致关键语义信息缺失,迫使模型只能依赖剩余的常用字进行对齐判别,从而引发预测偏差。最后,由于知识缺失引起语义关联失效,针对少部分古今异名的实体,模型仅凭字面和结构信息难以建立联系,这表明纯数据驱动的方法仍存在性能瓶颈。
本节应用GAFL-Align模型,实现《神农本草经》与《汤液本草》知识图谱的融合,并定量分析融合效果。构建过程采取分层混合对齐策略:针对“性味”“产地”“炮制方法”等语义明确且表述固定的实体,实施基于规则的精确匹配;对于“中药”“症状”“疾病”以及“功效”存在深层语义异构的复杂实体,部署GAFL-Align模型进行细粒度对齐。此外,为保障知识的完备性,两部典籍各自特有的知识均予以完整保留。最终,将所有被识别为等价的实体对合并为单一节点并迁移关联关系,生成融合图谱SKG-Fused。
图5所示,SKG-Fused在整体规模与拓扑复杂性上远超原始图谱,图右的“白薇”局部网络进一步展示了融合的具体价值。融合后的“白薇”节点汇集了来自两部典籍的知识,其关系网络得到了极大丰富,同时涵盖了功效、别名、产地以及采收方式等多个维度。这证明融合图谱成功构建了一个远比单一典籍更全面、更立体的实体知识网络。
SKG-Fused中“白薇”实体的知识来源构成如图6所示。针对“黄芪”与“黄耆”这对经典异名以及“寒热酸痛”与“寒热酸痋”这类存在古今字形差异的样本,模型利用BERT的深层语义编码与GAT的结构聚合能力,成功跨越了字面符号的差异并将其精准识别为等价实体。对于“卒中”和“神昏”这两类已标准化的实体,虽然其在两部典籍中名称一致,但各自关联的药物与症状网络存在显著异构性,GAFL-Align通过对抗学习消除分布偏差,并从结构层面识别出来源不同的两个节点实为同一对象从而实现了合并。这些关键实体在融合后被标记为同时来自两部典籍,这种基于模型计算的融合不仅实现了知识的统一,更通过多源记载的交叉验证显著增强了核心本草知识的可信度。
本节对融合图谱SKG-Fused与两个源图谱进行了定量指标对比,如表6所示。融合图谱在规模上接近源图谱总和,其中855对核心实体通过混合对齐策略被成功合并。分析网络指标可以发现,SKG-Fused的平均度达到了3.720 9,高于两个源图谱的3.642 4与3.371 4。这一结果证明了实体对齐的增益效应,即通过合并等价实体,融合后的节点继承了更丰富的关系属性从而充实了其知识内涵。同时,融合图谱的平均聚类系数为0.063 1,体现了对两个源图谱的有效中和与继承,说明融合策略能够将结构稀疏与紧密的图谱进行有机整合。值得注意的是,融合图谱的图密度有所下降,但这属于符合图论规律的预期现象。此外,SKG-Fused的最大连通分量占比高达98.71%,表明图谱的核心主体保持了高度连通。知识融合并未导致图结构的碎裂,而是在保持主体结构完整性的基础上提升了实体的平均知识丰裕度。
本研究在方法创新层面突破了本草典籍实体对齐主要依靠人工对齐与表层规则匹配的局限,构建了面向古籍复杂语义特性的自动化对齐框架。针对传统规则方法难以应对典籍异构与术语易混淆的特有挑战,本研究提出了GAFL-Align模型,通过生成对抗网络与模糊边界辨识机制有效攻克了深层语义识别难题。此外,针对古籍领域专家标注资源稀缺且昂贵的困境,本研究引入了置信度感知的迭代自训练策略,利用高置信度伪标签在保证精度的同时大幅降低了对大规模人工标注数据的依赖。本模型实现了低资源场景下的高精度、自动化对齐,为古籍知识的智能化组织提供了高效的方法论支撑。
在知识组织层面,本研究改善了既往本草知识图谱研究中构建与应用环节衔接不足的现状,确立了以融合为核心的知识增值路径。区别于现有研究多局限于单一典籍知识图谱的构建或局部应用开发,本研究将重心聚焦于连接异构图谱的语义桥梁的构建,通过实体对齐技术,实现了《神农本草经》与《汤液本草》典籍内容的对齐与融合。这一工作打通了从上游图谱构建到下游智能应用的数据路径,构建了更加完整有序的本草知识网络,提升了本草知识的可信度,并为后续典籍可视化分析、智能问答及辅助诊疗等下游任务提供更为完备的数据支持。
相比于大语言模型方案,GAFL-Align在处理效率上具备显著优势。作为专用小模型,其利用图注意力网络进行大规模并行计算,从而避免了通用大模型将对齐转化为序列化文本检索带来的高推理延迟,更适合海量典籍实体的快速处理。除了速度优势,本模型在可解释性上也提供了独特的算法视角。尽管大模型能生成直观的自然语言解释,但其内部逻辑仍具有一定的黑盒属性,而GAFL-Align能够通过自适应门控机制动态展示语义与结构特征的权重分配,体现决策时的侧重依据,提供了基于算法机制的透明度。更重要的是,本模型摆脱了对高性能计算集群的依赖,仅需单显卡即可高效运行。这种轻量化架构显著降低了硬件门槛与运行开销,为本草古籍数字化整理工作的低成本、规模化开展提供了更具可行性的技术路径。
尽管GAFL-Align模型在本草典籍实体对齐任务中表现优异,但仍存在以下局限。首先,典籍间的知识组织逻辑存在较大差异时,基于对抗学习的特征映射效果便会受到限制。其次,通用预训练语言模型难以全面覆盖专业领域生僻字,导致关键术语的核心语义信息表征缺失,模型被迫过度依赖邻域结构,可能会产生预测偏差。最后,对于字面完全不同的古今异名实体,纯数据驱动的方法不能利用外部背景知识建立有效的隐式关联。
未来工作将致力于扩展实验规模,引入更多典籍数据以验证模型在极端异构场景下的泛化能力;构建本草专用嵌入模型并引入汉字字形或偏旁部首等视觉特征,解决生僻字编码难题;探索轻量化模型与大语言模型的协同机制,为对齐结果提供直观的自然语言事后解释,实现算法透明度与用户可理解性的双重提升。
本研究聚焦于本草典籍知识图谱的实体对齐任务,提出了一种融合生成对抗网络与模糊语义辨识的实体对齐模型GAFL-Align。在方法论层面,本模型通过引入对抗性领域自适应机制,有效消除了不同典籍间的跨图谱分布偏差,实现了深层的语义对齐。同时,本研究设计了模糊边界负采样策略与置信度感知的迭代自训练机制,不仅攻克了形近义远术语的细粒度辨识难题,更验证了在低资源环境下降低对人工标注依赖、实现自动化知识聚合的可行性。实验结果表明,GAFL-Align在Hits@1与MRR等关键指标上均优于当前主流的基线方法,证明了其在处理垂直领域复杂异构数据方面的优势。在此基础上,本研究整合了《神农本草经》与《汤液本草》中分散的知识资源,构建了一个多源融合知识图谱。综上所述,本研究不仅提升了本草知识体系的完整性,更为本草典籍的学术溯源、系统性整理以及传统医学智慧的数字化活化提供了坚实的方法支撑与数据基础。
参考文献 引证文献
排序方式:
1
沈旺,刘嘉宇,李贺,.学科领域生命周期视域下国内中医古籍数字化研究进展[J].图书情报工作202266(22):4-15.
2
赵辉,张华敏,张磊,.中医古籍文献研究方法述要[J].中国中医药图书情报杂志202448(1):228-233.
3
谢云霏,贾李蓉.中医药知识图谱构建技术及应用的研究进展[J].中国医药导报202421(20):62-66.
4
王松,李正钧,杨涛,.中医药知识图谱研究现状及发展趋势[J].南京中医药大学学报202238(3):272-278.
5
Weng HChen J LOu A Het al.Leveraging Representation Learning for the Construction and Application of a Knowledge Graph for Traditional Chinese Medicine:Framework Development Study[J].JMIR Medical Informatics202210(9):e38414.
6
Sun Z QZhang Q HHu Wet al.A Benchmarking Study of Embedding-Based Entity Alignment for Knowledge Graphs[J].Proceedings of the VLDB Endowment202013(12):2326-2340.
7
翟东升,娄莹,阚慧敏,.基于多源异构数据的中医药知识图谱构建与应用研究[J].数据分析与知识发现20237(9):146-158.
8
Mao XWang W TXu H Met al.Relational Reflection Entity Alignment[C]//Proceedings of the 29th ACM International Conference on Information and Knowledge Management.New York:Association for Computing Machinery,2020:1095-1104.
9
Tang X BZhang JChen Bet al.BERT-INT:A BERT-Based Interaction Model for Knowledge Graph Alignment[C]//Bessiere C.Proceedings of the 29th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2020:3174-3180.
10
阮彤,孙程琳,王昊奋,.中医药知识图谱构建与应用[J].医学信息学杂志201637(4):8-13.
11
Chen M HTian Y TYang M Het al.Multilingual Know⁃ledge Graph Embeddings for Cross-Lingual Knowledge Alignment[C]//Sierra C.Proceedings of the 26th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2017:1511-1517.
12
Sun Z QHu WLi C K.Cross-Lingual Entity Alignment via Joint Attribute-Preserving Embedding[C]//DAmato C,Fernandez M,Tamma V,et al.Semantic Web-ISWC 2017.Cham:Springer,2017:628-644.
13
Zhang Q HSun Z QHu Wet al.Multi-View Knowledge Graph Embedding for Entity Alignment[C]//Kraus S.Proceedings of the 28th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2019:5429-5435.
14
杨一帆,刘忠博,周玉新.中医知识图谱构建及应用综述[J].软件导刊202524(8):17-26.
15
于彤,刘静,贾李蓉,.大型中医药知识图谱构建研究[J].中国数字医学201510(3):80-82.
16
Zhou YQi X LHuang Yet al.Research on Construction and Application of TCM Knowledge Graph Based on Ancient Chinese Texts[C]//Barnaghi P,Gottlob G,Katsaros D,et al.Proceedings of the 2019 IEEE International Conference on Bioinformatics and Biomedicine.Piscataway:IEEE,2019:144-147.
17
李芊芊,付兴,杨凤,.基于“病脉证并治”诊疗思维的《伤寒论》知识图谱构建与应用[J].世界科学技术—中医药现代化202224(9):3613-3621.
18
张卫东,张晓晓.中医古籍数字资源知识组织与可视化研究——以《金匮要略》为例[J].情报科学202240(8):107-117.
19
李贺,祝琳琳,刘嘉宇,.基于本体的简帛医药知识组织研究[J].图书情报工作202266(22):16-27.
20
时倩如,李贺,于雯倩,.中医文本命名实体识别研究综述[J].现代情报202545(2):4-16.
21
李明,罗晓兰,朱邦贤.中医古籍方剂数据挖掘与知识问答系统构建[J].图书馆论坛202545(4):49-59.
22
许天翼,石进.基于循环-过滤检索增强生成技术的中医知识问答系统研究[J/OL].数据分析与知识发现:1-18[2026-02-02].https://link.cnki.net/urlid/10.1478.G2.20250815.1141.006.
23
沈旺,于琳,冯欣,.基于动态知识图谱的中医疫病古籍知识演化研究[J].现代情报202545(2):26-37.
24
杨凤,张艺,陶晓华,.基于知识元理论与临床需求深度融合的中医古籍数智化研究路径和范式[J].中医杂志202465(12):1201-1207.
25
齐佳杰,魏东升,刘雨婷,.基于“病机链”理念的《脾胃论》湿热成痿肺金受邪论Neo4j知识图谱构建[J].中医药导报202430(11):172-177.
26
张君冬,杨松桦,严颖,.跨医学体系下医疗知识图谱的构建与药物预测研究——以动脉粥样硬化为例[J].情报理论与实践202447(2):178-188.
27
张富,杨琳艳,李健伟,.实体对齐研究综述[J].计算机学报202245(6):1195-1225.
28
Jiang WLiu Y NDeng X Y.Fuzzy Entity Alignment via Knowledge Embedding With Awareness of Uncertainty Measure[J].Neurocomputing2022468:97-110.
29
Wang C DWang W HLiang Q Yet al.Local and Global Structure-Aware Contrastive Framework for Entity Alignment[J].Neurocomputing2025624:129445.
30
Wang Z CLv Q SLan X Het al.Cross-Lingual Knowledge Graph Alignment via Graph Convolutional Networks[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing.Stroudsburg:Association for Computational Linguistics,2018:349-357.
31
Long M XWang J HMa J Xet al.LLM-Augmented Entity Alignment:An Unsupervised and Training-Free Framework[J].Neural Networks2026194:108139.
32
Pei S CYu LHoehndorf Ret al.Semi-Supervised Entity Alignment via Knowledge Graph Embedding With Awareness of Degree Difference[C]//Proceedings of the World Wide Web Conference.New York:Association for Computing Machinery,2019:3130-3136.
33
Li Y FChen LLiu C Fet al.Generative Adversarial Network for Unsupervised Multi-Lingual Knowledge Graph Entity Alignment[J].World Wide Web-Internet and Web Information Systems202326(5):2265-2290.
34
Mao XWang W TWu Y Bet al.Boosting the Speed of Entity Alignment 10x:Dual Attention Matching Network With Normalized Hard Sample Mining[C]//Proceedings of the 30th World Wide Web Conference.New York:Association for Computing Machinery,2021:821-832.
35
Meng X ZJiao X XLi J Het al.CELLMEA:A Collaboratively Enhanced Large Language Model-Based Entity Alignment for Aircraft Fault Maintenance[J].Expert Systems With Applications2025282:127630.
36
Sun Z QHu WZhang Q Het al.Bootstrapping Entity Alignment With Knowledge Graph Embedding[C]//Lang J.Proceedings of the 27th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2018:4396-4402.
37
Liu BLan T CHua Wet al.Dependency-Aware Self-Training for Entity Alignment[C]//Proceedings of the 16th ACM International Conference on Web Search and Data Mining.New York:Association for Computing Machinery,2023:796-804.
38
Wu Y TLiu XFeng Y Set al.Relation-Aware Entity Alignment for Heterogeneous Knowledge Graphs[C]//Kraus S.Proceedings of the 28th International Joint Conference on Artificial Intelligence.California:International Joint Conferences on Artificial Intelligence Organization,2019:5278-5284.
39
Cao Y XLiu Z YLi C Jet al.Multi-Channel Graph Neural Network for Entity Alignment[C]//Korhonen A,Traum D,Marquez L.Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.Stroudsburg:Association for Computational Linguistics,2019:1452-1461.
40
Zhang RSu Y XTrisedya B Det al.AutoAlign:Fully Automatic and Effective Knowledge Graph Alignment Enabled by Large Language Models[J].IEEE Transactions on Knowledge and Data Engineering202436(6):2357-2371.
41
Jiang X HShen Y HShi Z Cet al.Unlocking the Power of Large Language Models for Entity Alignment[C]//Ku L W,Martins A,Srikumar V.Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics.Stroudsburg:Association for Computational Linguistics,2024:7566-7583.
2026年第46卷第3期
PDF下载
27
12
引用本文
BibTeX
文章信息
doi: 10.3969/j.issn.1008-0821.2026.03.003
  • 接收时间:2025-12-11
  • 首发时间:2026-06-05
  • 出版时间:2026-03-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-12-11
基金
作者信息
    1吉林大学商学与管理学院,吉林长春130012
    2吉林省图书馆,吉林长春130022

通讯作者:

李贺(1964-),女,教授,博士,博士生导师,研究方向:信息行为分析、知识管理。
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/xdqb/CN/10.3969/j.issn.1008-0821.2026.03.003
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏