Article(id=1251458156577108696, tenantId=1146029695717560320, journalId=1251194880429441115, issueId=1251458153020342360, articleNumber=null, orderNo=null, doi=10.3979/j.issn.1673-825X.202408040201, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1722700800000, receivedDateStr=2024-08-04, revisedDate=1757088000000, revisedDateStr=2025-09-06, acceptedDate=null, acceptedDateStr=null, onlineDate=1776300475496, onlineDateStr=2026-04-16, pubDate=null, pubDateStr=null, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1776300475496, onlineIssueDateStr=2026-04-16, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1776300475496, creator=13041195026, updateTime=1776300475496, updator=13041195026, issue=Issue{id=1251458153020342360, tenantId=1146029695717560320, journalId=1251194880429441115, year='2025', volume='37', issue='5', pageStart='627', pageEnd='780', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=1, specialIssue=null, createTime=1776300474648, creator=13041195026, updateTime=1776311939434, updator=13041195026, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1251506239914586238, tenantId=1146029695717560320, journalId=1251194880429441115, issueId=1251458153020342360, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1251506239914586239, tenantId=1146029695717560320, journalId=1251194880429441115, issueId=1251458153020342360, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=708, endPage=716, ext={EN=ArticleExt(id=1251458156933624540, articleId=1251458156577108696, tenantId=1146029695717560320, journalId=1251194880429441115, language=EN, title=Temporal knowledge graph reasoning model incorporating cross-time commonality features, columnId=1251458154354131041, journalTitle=Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), columnName=Artificial Intelligenceand Big Data, runingTitle=null, highlight=null, articleAbstract=

Temporal knowledge graph reasoning, which predicts events absent from the graph, has seen significant applications in recommendation systems, question answering, and healthcare. The lack of background knowledge in temporal knowledge graphs hinders reasoning, with existing methods relying on external graphs while overlooking implicit data within the graph. To fully exploit the graph's implicit background information, this paper extracts cross-temporal features to define entity backgrounds and proposes a temporal knowledge graph reasoning model incorporating cross-time commonality features(TR-CTC). TR-CTC uses a graph neural network to extract cross-temporal commonality from multi-hop paths, integrating it as background information into the graph representation learning process, enhancing reasoning performance. Experimental results show that TR-CTC generally outperforms baseline models in link prediction tasks.

, correspAuthors=null, 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=Meiqi CHEN, Chenglin ZHANG, Hong YU), CN=ArticleExt(id=1251458175757660485, articleId=1251458156577108696, tenantId=1146029695717560320, journalId=1251194880429441115, language=CN, title=融合跨时间共性特征的时序知识图谱推理模型, columnId=1251458154492543075, journalTitle=重庆邮电大学学报(自然科学版), columnName=人工智能与大数据, runingTitle=null, highlight=null, articleAbstract=

时序知识图谱推理用于预测未出现在时序知识图谱中的事件信息,在推荐系统、问答系统和医疗等场景下取得了重要的应用。时序知识图谱中实体相关背景知识的缺失,限制了推理技术的发展。现有的推理方法依赖于外部图谱,却很少关注图谱自身所隐含的背景信息。为了充分挖掘图谱隐含的背景信息,通过抽取跨时间共性的特征刻画实体背景,提出了一种融合了跨时间共性特征的时序知识图谱推理模型(temporal knowledge graph reasoning model incorporating cross-time commonality features,TR-CTC)。TR-CTC利用图神经网络模型,从多跳路径中抽取跨时间共性特征作为实体的背景信息;融入到时序知识图谱表示学习过程中,提升推理性能。实验结果表明,TRCTC在链接预测任务中普遍优于基准模型。

, correspAuthors=null, authorNote=null, correspAuthorsNote=
于洪
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=r4hHjSy6F2/is0Kbtwai9A==, magXml=w1O2+uBwXGgAc3hect1+rg==, pdfUrl=null, pdf=dNhaPiQtPJNMbNCjJd7sYA==, pdfFileSize=8462383, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=mpQchZz8YcygpA225x5cTA==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=Tijvi8315WVQaN2ehq7wxQ==, mapNumber=null, authorCompany=null, fund=null, authors=

陈美琪,硕士研究生,主要研究方向为知识图谱和知识表示。E-mail:

张诚麟,博士研究生,主要研究方向为因果学习和知识表示。E-mail:

于洪,教授,博士,主要研究方向为粗糙集、粒计算、三支决策、智能信息处理、Web智能、数据挖掘。E-mail:

, authorsList=陈美琪, 张诚麟, 于洪)}, authors=[Author(id=1251458176089010527, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=s220201012@stu.cqupt.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1251458176156119397, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, authorId=1251458176089010527, language=EN, stringName=Meiqi CHEN, firstName=Meiqi, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251458176240005484, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, authorId=1251458176089010527, language=CN, stringName=陈美琪, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=重庆邮电大学 计算智能重庆市重点实验室,重庆 400065, bio={"content":"

陈美琪,硕士研究生,主要研究方向为知识图谱和知识表示。E-mail:

"}, bioImg=null, bioContent=

陈美琪,硕士研究生,主要研究方向为知识图谱和知识表示。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251458175984152919, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, xref=null, ext=[AuthorCompanyExt(id=1251458175992541528, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China), AuthorCompanyExt(id=1251458175996735833, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=重庆邮电大学 计算智能重庆市重点实验室,重庆 400065)])]), Author(id=1251458176298725744, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=d210201035@stu.cqupt.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1251458176391000439, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, authorId=1251458176298725744, language=EN, stringName=Chenglin ZHANG, firstName=Chenglin, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251458176462303611, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, authorId=1251458176298725744, language=CN, stringName=张诚麟, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=重庆邮电大学 计算智能重庆市重点实验室,重庆 400065, bio={"content":"

张诚麟,博士研究生,主要研究方向为因果学习和知识表示。E-mail:

"}, bioImg=null, bioContent=

张诚麟,博士研究生,主要研究方向为因果学习和知识表示。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251458175984152919, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, xref=null, ext=[AuthorCompanyExt(id=1251458175992541528, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China), AuthorCompanyExt(id=1251458175996735833, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=重庆邮电大学 计算智能重庆市重点实验室,重庆 400065)])]), Author(id=1251458176554578306, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=yuhong@cqupt.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1251458176630075781, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, authorId=1251458176554578306, language=EN, stringName=Hong YU, firstName=Hong, middleName=null, lastName=YU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1251458176718156171, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, authorId=1251458176554578306, language=CN, stringName=于洪, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=重庆邮电大学 计算智能重庆市重点实验室,重庆 400065, bio={"content":"

于洪,教授,博士,主要研究方向为粗糙集、粒计算、三支决策、智能信息处理、Web智能、数据挖掘。E-mail:

"}, bioImg=null, bioContent=

于洪,教授,博士,主要研究方向为粗糙集、粒计算、三支决策、智能信息处理、Web智能、数据挖掘。E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1251458175984152919, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, xref=null, ext=[AuthorCompanyExt(id=1251458175992541528, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China), AuthorCompanyExt(id=1251458175996735833, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=重庆邮电大学 计算智能重庆市重点实验室,重庆 400065)])])], keywords=[Keyword(id=1251458176839790991, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, orderNo=1, keyword=temporal knowledge graph), Keyword(id=1251458176902705554, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, orderNo=2, keyword=graph neural network), Keyword(id=1251458176961425814, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, orderNo=3, keyword=link prediction), Keyword(id=1251458177041117595, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, orderNo=4, keyword=cross-temporal commonality), Keyword(id=1251458178643341727, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, orderNo=5, keyword=knowledge graph reasoning), Keyword(id=1251458178739810721, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, orderNo=1, keyword=时序知识图谱), Keyword(id=1251458178907582886, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, orderNo=2, keyword=图神经网络), Keyword(id=1251458179100520872, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, orderNo=3, keyword=链接预测), Keyword(id=1251458179247321514, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, orderNo=4, keyword=跨时间共性), Keyword(id=1251458179347984814, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, orderNo=5, keyword=知识图谱推理)], refs=[Reference(id=1251458181197672967, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2022, volume=41, issue=2, pageStart=1, pageEnd=26, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=ZHAO Y, WANG X, CHEN J, journalName=ACM Transactions on Information Systems, refType=null, unstructuredReference=ZHAO Y, WANG X, CHEN J, et al. Time-aware path reasoning on knowledge graph for recommendation[J]. ACM Transactions on Information Systems, 2022, 41(2): 1-26., articleTitle=Time-aware path reasoning on knowledge graph for recommendation, refAbstract=null), Reference(id=1251458181277364746, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=5825, pageEnd=5833, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=MAVROMATIS C, SUBRAMANYAM P L, IOANNIDIS V N, journalName=null, refType=null, unstructuredReference=MAVROMATIS C, SUBRAMANYAM P L, IOANNIDIS V N, et al. Tempoqr: Temporal question reasonning over knowledge graphs[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Washington, USA: AAAI, 2022: 5825-5833., articleTitle=Tempoqr: Temporal question reasonning over knowledge graphs, refAbstract=null), Reference(id=1251458181357056526, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2017, volume=54, issue=12, pageStart=2687, pageEnd=2697, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=陈德华, 殷苏娜, 乐嘉锦, journalName=计算机研究与发展, refType=null, unstructuredReference=陈德华,殷苏娜,乐嘉锦,.一种面向临床领域时序知识图谱的链接预测模型[J].计算机研究与发展, 2017, 54(12): 2687-2697., articleTitle=一种面向临床领域时序知识图谱的链接预测模型, refAbstract=null), Reference(id=1251458181424165394, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2017, volume=54, issue=12, pageStart=2687, pageEnd=2697, url=null, language=null, rfNumber=[3], rfOrder=3, authorNames=CHENDH, YIN S N, LE J J, journalName=Journal of Computer Research and Development, refType=null, unstructuredReference=CHENDH,YIN S N, LE J J, et al. A link prediction model for clinical temporal knowledge graph[J]. Journal of Computer Research and Development, 2017, 54(12):2687-2697., articleTitle=A link prediction model for clinical temporal knowledge graph, refAbstract=null), Reference(id=1251458181499662871, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2008, volume=null, issue=null, pageStart=1247, pageEnd=1250, url=null, language=null, rfNumber=[4], rfOrder=4, authorNames=BOLLACKER K, EVANS C, PARITOSH P, journalName=null, refType=null, unstructuredReference=BOLLACKER K, EVANS C, PARITOSH P, et al. Free base: A collaboratively created graph database for structuring human knowledge[C]//Proceedings of the 2008 ACM SIGMOD International Conference on Management of Data. New York, USA: ACM2008: 1247-1250., articleTitle=Free base: A collaboratively created graph database for structuring human knowledge, refAbstract=null), Reference(id=1251458183110275610, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2007, volume=null, issue=null, pageStart=697, pageEnd=706, url=null, language=null, rfNumber=[5], rfOrder=5, authorNames=FABIAN M, GJERGJI K, GERHARD W, journalName=null, refType=null, unstructuredReference=FABIAN M, GJERGJI K, GERHARD W. Yago: A core of semantic knowledge unifying wordnet and wikipedia[C]//16th International World Wide Web Conference. Massachusetts, USA: WWW. 2007: 697-706., articleTitle=Yago: A core of semantic knowledge unifying wordnet and wikipedia, refAbstract=null), Reference(id=1251458183198355997, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=1, pageEnd=5, url=null, language=null, rfNumber=[6], rfOrder=6, authorNames=SUN Z, DENG Z H, NIE J Y, journalName=null, refType=null, unstructuredReference=SUN Z, DENG Z H, NIE J Y, et al. RotatE: Knowledge graph embedding by relational rotation in complex space[C]//International Conference on Learning Representations. Vancouver, Canada: ICLR, 2018: 1-5., articleTitle=RotatE: Knowledge graph embedding by relational rotation in complex space, refAbstract=null), Reference(id=1251458183269659167, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=32, pageEnd=40, url=null, language=null, rfNumber=[7], rfOrder=7, authorNames=DETTMERS T, MINERVINI P, STENETORP P, journalName=null, refType=null, unstructuredReference=DETTMERS T, MINERVINI P, STENETORP P, et al. Convolutional 2d knowledge graph embeddings[C]//Proceedings of the AAAI Conference on Artificial Intelligence.Washington, USA: AAAI, 2018: 32-40., articleTitle=Convolutional 2d knowledge graph embeddings, refAbstract=null), Reference(id=1251458183382905378, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2024, volume=35, issue=8, pageStart=3923, pageEnd=3951, url=null, language=null, rfNumber=[8], rfOrder=8, authorNames=王俞涵, 陈子阳, 赵翔, journalName=软件学报, refType=null, unstructuredReference=王俞涵,陈子阳,赵翔,.时序知识图谱表示与推理的研究进展与趋势[J].软件学报, 2024, 35(8): 3923-3951., articleTitle=时序知识图谱表示与推理的研究进展与趋势, refAbstract=null), Reference(id=1251458183466791460, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2024, volume=35, issue=8, pageStart=3923, pageEnd=3951, url=null, language=null, rfNumber=[8], rfOrder=9, authorNames=WANG Y H, CHEN Z Y, ZHAO X, journalName=Journal of Software, refType=null, unstructuredReference=WANG Y H, CHEN Z Y, ZHAO X, et al. Research progress and trend of temporal knowledge graph representation and reasoning[J]. Journal of Software, 2024, 35(8): 3923-3951., articleTitle=Research progress and trend of temporal knowledge graph representation and reasoning, refAbstract=null), Reference(id=1251458183550677543, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2016, volume=null, issue=null, pageStart=2350, pageEnd=2354, url=null, language=null, rfNumber=[9], rfOrder=10, authorNames=JIANG T, LIU T, GE T, journalName=null, refType=null, unstructuredReference=JIANG T, LIU T, GE T, et al. Encoding temporal information for time-aware link prediction[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language processing. Austin, Texas, USA: EMNLP, 2016: 2350-2354., articleTitle=Encoding temporal information for time-aware link prediction, refAbstract=null), Reference(id=1251458183630369323, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=4816, pageEnd=4821, url=null, language=null, rfNumber=[10], rfOrder=11, authorNames=GARCÍA-DURÁN A, DUMANCIC S, NIEPERT M, journalName=null, refType=null, unstructuredReference=GARCÍA-DURÁN A, DUMANCIC S, NIEPERT M. Learning sequence encoders for temporal knowledge graph completion[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. Brussels, Belgium: EMNLP, 2018: 4816-4821., articleTitle=Learning sequence encoders for temporal knowledge graph completion, refAbstract=null), Reference(id=1251458183714255404, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=4289, pageEnd=4300, url=null, language=null, rfNumber=[11], rfOrder=12, authorNames=KAZEMI S M, POOLE D, journalName=null, refType=null, unstructuredReference=KAZEMI S M, POOLE D. SimplE embedding for link prediction in knowledge graphs[C]//Proceedings of the International Conference on Neural Information Processing Systems. Siem Reap, Cambodia: ICONIP, 2018: 4289-4300., articleTitle=SimplE embedding for link prediction in knowledge graphs, refAbstract=null), Reference(id=1251458183798141486, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=6471, pageEnd=6479, url=null, language=null, rfNumber=[12], rfOrder=13, authorNames=SADEGHIAN A, ARMANDPOUR M, COLAS A, journalName=null, refType=null, unstructuredReference=SADEGHIAN A, ARMANDPOUR M, COLAS A, et al. Chronor: Rotation based temporal knowledge graph embedding[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Washington, USA: AAAI, 2021:6471-6479., articleTitle=Chronor: Rotation based temporal knowledge graph embedding, refAbstract=null), Reference(id=1251458183877833263, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=2001, pageEnd=2011, url=null, language=null, rfNumber=[13], rfOrder=14, authorNames=DASGUPTA S S, RAY S N, TALUKDAR P, journalName=null, refType=null, unstructuredReference=DASGUPTA S S, RAY S N, TALUKDAR P. Hyte: Hyperplane-based temporally aware knowledge graph embedding[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural: Language Processing. Brussels, Belgium: EMNLP, 2018: 2001-2011., articleTitle=Hyte: Hyperplane-based temporally aware knowledge graph embedding, refAbstract=null), Reference(id=1251458183961719349, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2015, volume=null, issue=null, pageStart=3920, pageEnd=3927, url=null, language=null, rfNumber=[14], rfOrder=15, authorNames=MUTHIAH S, HUANG B, ARREDONDO J, journalName=null, refType=null, unstructuredReference=MUTHIAH S, HUANG B, ARREDONDO J, et al. Planned protest modeling in news and social media[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Austin, Texas, USA: AAAI, 2015:3920-3927., articleTitle=Planned protest modeling in news and social media, refAbstract=null), Reference(id=1251458184062382649, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=6669, pageEnd=6683, url=null, language=null, rfNumber=[15], rfOrder=16, authorNames=JIN W, QU M, JIN X, journalName=null, refType=null, unstructuredReference=JIN W, QU M, JIN X, et al. Recurrent event network:Autoregressive structure inference over temporal knowledge graphs[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. Punta Cana, Dominican Republic: EMNLP, 2020: 6669-6683., articleTitle=Recurrent event network:Autoregressive structure inference over temporal knowledge graphs, refAbstract=null), Reference(id=1251458184146268735, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=593, pageEnd=607, url=null, language=null, rfNumber=[16], rfOrder=17, authorNames=SCHLICHTKRULL M, KIPF T N, BLOEM P, journalName=null, refType=null, unstructuredReference=SCHLICHTKRULL M, KIPF T N, BLOEM P, et al. Modeling relational data with graph convolutional networks[C]//The Semantic Web: 15th International Conference. Heraklion Crete, Greece: Springer, 2018: 593-607., articleTitle=Modeling relational data with graph convolutional networks, refAbstract=null), Reference(id=1251458184242737730, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2014, volume=null, issue=null, pageStart=1724, pageEnd=1734, url=null, language=null, rfNumber=[17], rfOrder=18, authorNames=CHO K, MERRIENBOER B V, GÜLÇEHREÇ, journalName=null, refType=null, unstructuredReference=CHO K, MERRIENBOER B V, GÜLÇEHREÇ, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. Doha, Qatar: EMNLP, 2014:1724-1734., articleTitle=Learning phrase representations using RNN encoder-decoder for statistical machine translation, refAbstract=null), Reference(id=1251458184355983941, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=794, pageEnd=803, url=null, language=null, rfNumber=[18], rfOrder=19, authorNames=PARK N, LIU F, MEHTA P, journalName=null, refType=null, unstructuredReference=PARK N, LIU F, MEHTA P, et al. EvoKG: Jointly modeling event time and network structure for reasoning over temporal knowledge graphs[C]//Proceedings of the 15th ACM International Conference on Web Search and Data Mining. New York, USA: ACM, 2022: 794-803., articleTitle=EvoKG: Jointly modeling event time and network structure for reasoning over temporal knowledge graphs, refAbstract=null), Reference(id=1251458184444064329, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=5730, pageEnd=5746, url=null, language=null, rfNumber=[19], rfOrder=20, authorNames=WU J, CAO M, CHEUNG J C K, journalName=null, refType=null, unstructuredReference=WU J, CAO M, CHEUNG J C K, et al. TeMP: Temporal message passing for temporal knowledge graph completion[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. Vienna, Austria: EMNLP, 2020: 5730-5746., articleTitle=TeMP: Temporal message passing for temporal knowledge graph completion, refAbstract=null), Reference(id=1251458184527950414, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=2152, pageEnd=2158, url=null, language=null, rfNumber=[20], rfOrder=21, authorNames=LI Y, SUN S, ZHAO J, journalName=null, refType=null, unstructuredReference=LI Y, SUN S, ZHAO J. TiRGN: Time-guided recurrent graph network with local-global historical patterns for temporal knowledge graph reasoning[C]//Proceedings of the 31th International Joint Conference on Artificial Intelligence. Vienna, Austria: IJCAI, 2022: 2152-2158., articleTitle=TiRGN: Time-guided recurrent graph network with local-global historical patterns for temporal knowledge graph reasoning, refAbstract=null), Reference(id=1251458184607642193, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2021, volume=null, issue=null, pageStart=408, pageEnd=417, url=null, language=null, rfNumber=[21], rfOrder=22, authorNames=LI Z, JIN X, LI W, journalName=null, refType=null, unstructuredReference=LI Z, JIN X, LI W, et al. Temporal knowledge graph reasoning based on evolutional representation learning[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2021: 408-417., articleTitle=Temporal knowledge graph reasoning based on evolutional representation learning, refAbstract=null), Reference(id=1251458184712499799, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2022, volume=null, issue=null, pageStart=7328, pageEnd=7338, url=null, language=null, rfNumber=[22], rfOrder=23, authorNames=LI Z, HOU Z, GUAN S, journalName=null, refType=null, unstructuredReference=LI Z, HOU Z, GUAN S, et al. HiSMatch: Historical structure matching based temporal knowledge graph reasoning[C]//Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. Abu Dhabi, United Arab Emirates: EMNLP, 2022: 7328-7338., articleTitle=HiSMatch: Historical structure matching based temporal knowledge graph reasoning, refAbstract=null), Reference(id=1251458184779608665, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=19, pageEnd=25, url=null, language=null, rfNumber=[23], rfOrder=24, authorNames=VASHISHTH S, SANYAL S, NITIN V, journalName=null, refType=null, unstructuredReference=VASHISHTH S, SANYAL S, NITIN V, et al. Composition-based multi-relational graph convolutional networks[C]//International Conference on Learning Representations. Addis Ababa, Ethiopia: ICLR, 2020: 19-25., articleTitle=Composition-based multi-relational graph convolutional networks, refAbstract=null), Reference(id=1251458184855106141, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=3060, pageEnd=3067, url=null, language=null, rfNumber=[24], rfOrder=25, authorNames=SHANG C, TANG Y, HUANG J, journalName=null, refType=null, unstructuredReference=SHANG C, TANG Y, HUANG J, et al. End-to-end structure-aware convolutional networks for knowledge base completion[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Honolulu, Hawaii, USA: AAAI, 2019: 3060-3067., articleTitle=End-to-end structure-aware convolutional networks for knowledge base completion, refAbstract=null), Reference(id=1251458184926409313, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2018, volume=null, issue=null, pageStart=969, pageEnd=976, url=null, language=null, rfNumber=[25], rfOrder=26, authorNames=NGUYEN G H, LEE J B, ROSSI R A, journalName=null, refType=null, unstructuredReference=NGUYEN G H, LEE J B, ROSSI R A, et al. Continuoustime dynamic network embeddings[C]//Proceedings of the 2018 World Wide Web Conference. Lyon, France:ACM, 2018:969-976., articleTitle=Continuoustime dynamic network embeddings, refAbstract=null), Reference(id=1251458185010295396, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2024-08-04, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[26], rfOrder=27, authorNames=LAUTENSCHLAGER J, SHELLMAN S, WARD M, journalName=null, refType=null, unstructuredReference=LAUTENSCHLAGER J, SHELLMAN S, WARD M. ICEWS events and aggregations[EB/OL].[2024-08-04]. https://data.nber.org/ens/feldstein/NBER_Sources/ENS%20Conference%20Sources/2016/Predicting%20Conflict%20Via%20Machine%20Learning/ICEWS%20(Lockheed%20Martin)/Dictionaries/ICEWS%20Events%20and%20Aggregations.pdf., articleTitle=ICEWS events and aggregations, refAbstract=null), Reference(id=1251458185094181479, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, doi=null, pmid=null, pmcid=null, year=2023, volume=null, issue=null, pageStart=566, pageEnd=576, url=null, language=null, rfNumber=[27], rfOrder=28, authorNames=GIRALDO J H, SKIANIS K, BOUWMANS T, journalName=null, refType=null, unstructuredReference=GIRALDO J H, SKIANIS K, BOUWMANS T, et al. On the trade-off between over-smoothing and over-squashing in deep graph neural networks[C]//Proceedings of the 32nd ACM International Conference on Information and knowledge management. Birmingham, United Kingdom:ACM, 2023: 566-576., articleTitle=On the trade-off between over-smoothing and over-squashing in deep graph neural networks, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1251458175984152919, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, xref=null, ext=[AuthorCompanyExt(id=1251458175992541528, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China), AuthorCompanyExt(id=1251458175996735833, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, companyId=1251458175984152919, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=重庆邮电大学 计算智能重庆市重点实验室,重庆 400065)])], figs=[ArticleFig(id=1251458179503174067, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Fig.1, caption=Examples of ICEWS14 dataset, figureFileSmall=MWZs13cVfPV96urLBrSIkQ==, figureFileBig=mpQchZz8YcygpA225x5cTA==, tableContent=null), ArticleFig(id=1251458179574477242, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=图1, caption=ICEWS14数据集中的实例, figureFileSmall=MWZs13cVfPV96urLBrSIkQ==, figureFileBig=mpQchZz8YcygpA225x5cTA==, tableContent=null), ArticleFig(id=1251458179754832326, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Fig.2, caption=Framework of TR-CTC, figureFileSmall=csyW/kWhdae/R3atkrJdag==, figureFileBig=uRxxnMiOuQpInnmjnKSX5A==, tableContent=null), ArticleFig(id=1251458179847107017, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=图2, caption=TR-CTC框架图, figureFileSmall=csyW/kWhdae/R3atkrJdag==, figureFileBig=uRxxnMiOuQpInnmjnKSX5A==, tableContent=null), ArticleFig(id=1251458179947770320, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Fig.3, caption=Results of λ in entity prediction on ICEWS14, figureFileSmall=Iv2bxN/bEKZGola4Jqrudw==, figureFileBig=Et0hte+ryouWH3v/5K/DXA==, tableContent=null), ArticleFig(id=1251458180014879186, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=图3, caption=λ在ICEWS14数据集上实体预测的结果, figureFileSmall=Iv2bxN/bEKZGola4Jqrudw==, figureFileBig=Et0hte+ryouWH3v/5K/DXA==, tableContent=null), ArticleFig(id=1251458180094570965, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Fig.4, caption=Results of λ in relation prediction on ICEWS14, figureFileSmall=YtJUVjMJDz5EWURQo8GdBA==, figureFileBig=WMV5necS8DjTSXfFJEq3lg==, tableContent=null), ArticleFig(id=1251458180161679831, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=图4, caption=λ在ICEWS14数据集上关系预测的结果, figureFileSmall=YtJUVjMJDz5EWURQo8GdBA==, figureFileBig=WMV5necS8DjTSXfFJEq3lg==, tableContent=null), ArticleFig(id=1251458180249760220, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Tab.1, caption=

Datasets of temporal knowledge graph

, figureFileSmall=null, figureFileBig=null, tableContent=
数据集ICEWS14ICEWS18ICEWS0515
实体个数71282303310488
关系个数230256251
训练集数量63685539286272115
验证集数量138236753817535
测试集数量132226311020466
), ArticleFig(id=1251458180325257697, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=表1, caption=

时序知识图谱数据集

, figureFileSmall=null, figureFileBig=null, tableContent=
数据集ICEWS14ICEWS18ICEWS0515
实体个数71282303310488
关系个数230256251
训练集数量63685539286272115
验证集数量138236753817535
测试集数量132226311020466
), ArticleFig(id=1251458180409143782, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Tab.2, caption=

Results of entity prediction

, figureFileSmall=null, figureFileBig=null, tableContent=
模型ICEWS14ICEWS18ICEWS0515
MRRH@1H@3H@10MRRH@1H@3H@10MRRH@1H@3H@10
TTransE[9]13.433.1117.3234.5510.174.5210.3321.2515.715.0019.7238.02
DE-DisMult[10]27.9119.9930.9543.1319.6311.8022.0535.2834.7025.2938.9958.35
DE-SimplE[11]32.7424.3736.2848.8022.6914.2225.6139.5735.0325.9438.8952.85
RE-Net[15]35.1125.8139.2653.2026.6516.9630.4045.7335.9126.0440.3455.09
RE-GCN[21]40.0230.3644.5758.6328.9418.9432.9348.5645.9135.4151.5265.98
EvoKG[18]27.1816.6726.6342.7026.8126.8933.9450.0938.2328.1643.0657.40
TR-CTC40.9731.3245.5359.4532.4222.4436.6151.8747.6037.1753.1067.70
), ArticleFig(id=1251458180509807084, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=表2, caption=

实体预测结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型ICEWS14ICEWS18ICEWS0515
MRRH@1H@3H@10MRRH@1H@3H@10MRRH@1H@3H@10
TTransE[9]13.433.1117.3234.5510.174.5210.3321.2515.715.0019.7238.02
DE-DisMult[10]27.9119.9930.9543.1319.6311.8022.0535.2834.7025.2938.9958.35
DE-SimplE[11]32.7424.3736.2848.8022.6914.2225.6139.5735.0325.9438.8952.85
RE-Net[15]35.1125.8139.2653.2026.6516.9630.4045.7335.9126.0440.3455.09
RE-GCN[21]40.0230.3644.5758.6328.9418.9432.9348.5645.9135.4151.5265.98
EvoKG[18]27.1816.6726.6342.7026.8126.8933.9450.0938.2328.1643.0657.40
TR-CTC40.9731.3245.5359.4532.4222.4436.6151.8747.6037.1753.1067.70
), ArticleFig(id=1251458180585304562, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Tab.3, caption=

Results of relation prediction

, figureFileSmall=null, figureFileBig=null, tableContent=
模型ICEWS14ICEWS18ICEWS0515
MRRH@1H@3H@10MRRH@1H@3H@10MRRH@1H@3H@10
TTransE[9]1.556.7025.902.666.4019.401.174.7318.82
DE-DisMult[10]17.5211.9019.6029.4713.588.8014.9923.325.932.746.2011.63
DE-SimplE[11]19.6314.2121.5331.2716.4111.1518.6126.7214.8110.1916.5123.98
RE-Net[15]35.1117.5130.5053.1026.6523.8344.3076.1635.9125.0052.2779.55
RE-GCN[21]43.6031.4647.8870.3639.6224.7346.0170.9240.0824.7646.9971.82
EvoKG[18]41.4725.6747.7775.9541.1524.9248.0976.7639.2123.3344.7275.67
TR-CTC43.9831.5848.9172.7044.2231.5648.9872.1445.5533.0550.6273.07
), ArticleFig(id=1251458180685967862, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=表3, caption=

关系预测结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型ICEWS14ICEWS18ICEWS0515
MRRH@1H@3H@10MRRH@1H@3H@10MRRH@1H@3H@10
TTransE[9]1.556.7025.902.666.4019.401.174.7318.82
DE-DisMult[10]17.5211.9019.6029.4713.588.8014.9923.325.932.746.2011.63
DE-SimplE[11]19.6314.2121.5331.2716.4111.1518.6126.7214.8110.1916.5123.98
RE-Net[15]35.1117.5130.5053.1026.6523.8344.3076.1635.9125.0052.2779.55
RE-GCN[21]43.6031.4647.8870.3639.6224.7346.0170.9240.0824.7646.9971.82
EvoKG[18]41.4725.6747.7775.9541.1524.9248.0976.7639.2123.3344.7275.67
TR-CTC43.9831.5848.9172.7044.2231.5648.9872.1445.5533.0550.6273.07
), ArticleFig(id=1251458180786631159, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Tab.4, caption=

Ablation experiment on link prediction

, figureFileSmall=null, figureFileBig=null, tableContent=
任务模型ICEWS14ICEWS18ICEWS0515
MRRH@1H@3H@10MRRH@1H@3H@10MRRH@1H@3H@10
实体w/o CTC36.5327.5240.0854.6728.1818.2533.0347.9647.1736.8752.5967.05
预测TR-CTC40.9731.3245.5359.4532.4222.4436.6151.8747.6037.1753.1067.70
关系w/o CTC41.8629.7745.4969.3038.0724.0243.3868.3044.4031.7049.2572.77
预测TR-CTC43.9831.5848.9172.7044.2231.5648.9872.1445.5533.0550.6273.07
), ArticleFig(id=1251458180908265980, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=表4, caption=

链接预测上的消融实验

, figureFileSmall=null, figureFileBig=null, tableContent=
任务模型ICEWS14ICEWS18ICEWS0515
MRRH@1H@3H@10MRRH@1H@3H@10MRRH@1H@3H@10
实体w/o CTC36.5327.5240.0854.6728.1818.2533.0347.9647.1736.8752.5967.05
预测TR-CTC40.9731.3245.5359.4532.4222.4436.6151.8747.6037.1753.1067.70
关系w/o CTC41.8629.7745.4969.3038.0724.0243.3868.3044.4031.7049.2572.77
预测TR-CTC43.9831.5848.9172.7044.2231.5648.9872.1445.5533.0550.6273.07
), ArticleFig(id=1251458181034095101, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=EN, label=Tab.5, caption=

Cross-temporal commonality analysis on ICEWS14

, figureFileSmall=null, figureFileBig=null, tableContent=
任务模型MRRH@1H@3H@10
实体预测TR-CTC(random)26.1618.6928.9139.80
TR-CTC(cos-sim)36.8327.3541.1155.26
TR-CTC40.9731.3245.5359.45
关系预测TR-CTC(random)42.2829.1847.0870.19
TR-CTC(cos-sim)41.8229.4845.6569.27
TR-CTC43.9831.5848.9172.70
), ArticleFig(id=1251458181101203970, tenantId=1146029695717560320, journalId=1251194880429441115, articleId=1251458156577108696, language=CN, label=表5, caption=

在ICEWS14上的跨时间共性有效性分析

, figureFileSmall=null, figureFileBig=null, tableContent=
任务模型MRRH@1H@3H@10
实体预测TR-CTC(random)26.1618.6928.9139.80
TR-CTC(cos-sim)36.8327.3541.1155.26
TR-CTC40.9731.3245.5359.45
关系预测TR-CTC(random)42.2829.1847.0870.19
TR-CTC(cos-sim)41.8229.4845.6569.27
TR-CTC43.9831.5848.9172.70
)], attaches=null, journal=Journal(id=1251194162217791579, delFlag=0, nameCn=重庆邮电大学学报(自然科学版), nameEn=Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), nameHistory1=null, nameHistory2=null, issn=1673-825X, eissn=null, cn=50-1181/N, coden=null, periodic=1, 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=eIjswk9Qxcq7+V27dEZ90g==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), journalRemark=null, publicationField=null, createdTime=1776237534337, updatedTime=1776238167705, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Natural Sciences, subjectName=null, subjectCodeEn=Natural Sciences, subjectNameEn=null, picCn=eIjswk9Qxcq7+V27dEZ90g==, picEn=PN1NU8XEwLhVlBzwQEoroA==, jcr=null, cjcr=null, exts=[JournalExt(id=1251196818869137642, language=CN, name=重庆邮电大学学报(自然科学版), 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=1776238167732, updatedTime=1776238167732, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://journal2.cqupt.edu.cn/jcuptnsecn/jcuptnse/author/login, submissionEditorUrl=http://journal2.cqupt.edu.cn/jcuptnsecn/jcuptnse/editor/login, submissionReviewUrl=http://journal2.cqupt.edu.cn/jcuptnsecn/jcuptnse/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1251196818936246507, language=EN, name=Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), 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=1776238167747, updatedTime=1776238167747, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=http://journal2.cqupt.edu.cn/jcuptnsecn/jcuptnse/author/login, submissionEditorUrl=http://journal2.cqupt.edu.cn/jcuptnsecn/jcuptnse/editor/login, submissionReviewUrl=http://journal2.cqupt.edu.cn/jcuptnsecn/jcuptnse/reviewer/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1251194880429441115, websiteList=[Website(id=1251197148419797361, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251194880429441115, 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/cqyddxxb/CN, language=CN, createTime=1776238246302, createBy=18614031015, updateTime=1776238652182, updateBy=18614031015, name=重庆邮电大学学报(自然科学版)-中文, tplId=1146099689490845704, title=重庆邮电大学学报(自然科学版), delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251198981565530670, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=articleTextType, value=kx, createTime=1776238683358, updateTime=1776238683358, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981540364843, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=banner, value=null, createTime=1776238683352, updateTime=1776238683352, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981590696497, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=grayFlag, value=0, createTime=1776238683364, updateTime=1776238683364, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981531976234, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=logo, value=https://castjournals.cast.org.cn/joweb/cqyddxxb/CN/file/pic?fileId=X24gYotabwCg03WC1YllbA==, createTime=1776238683350, updateTime=1776238683350, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981607473715, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=minRunFlag, value=0, createTime=1776238683368, updateTime=1776238683368, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981557142061, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/cqyddxxb/CN/file/pic, createTime=1776238683356, updateTime=1776238683356, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981599085106, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=silenceFlag, value=0, createTime=1776238683366, updateTime=1776238683366, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981548753452, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1776238683354, updateTime=1776238683354, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981578113583, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=themeColor, value=null, createTime=1776238683361, updateTime=1776238683361, creator=18614031015, updator=18614031015), WebsiteProps(id=1251198981582307888, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148419797361, code=themeStyle, value=null, createTime=1776238683362, updateTime=1776238683362, creator=18614031015, updator=18614031015)]), Website(id=1251197148516266372, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1251194880429441115, 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/cqyddxxb/EN, language=EN, createTime=1776238246325, createBy=18614031015, updateTime=1776238647495, updateBy=18614031015, name=重庆邮电大学学报(自然科学版)-英文, tplId=1146101810881728533, title=Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), delFlag=0, indexPage=/home, props=[WebsiteProps(id=1251199007897371192, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=articleTextType, value=kx, createTime=1776238689636, updateTime=1776238689636, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007872205365, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=banner, value=null, createTime=1776238689630, updateTime=1776238689630, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007914148411, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=grayFlag, value=0, createTime=1776238689640, updateTime=1776238689640, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007863816756, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=logo, value=https://castjournals.cast.org.cn/joweb/cqyddxxb/EN/file/pic?fileId=X24gYotabwCg03WC1YllbA==, createTime=1776238689628, updateTime=1776238689628, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007930925629, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=minRunFlag, value=0, createTime=1776238689644, updateTime=1776238689644, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007888982583, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/cqyddxxb/EN/file/pic, createTime=1776238689634, updateTime=1776238689634, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007922537020, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=silenceFlag, value=0, createTime=1776238689642, updateTime=1776238689642, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007880593974, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1776238689632, updateTime=1776238689632, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007901565497, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=themeColor, value=null, createTime=1776238689637, updateTime=1776238689637, creator=18614031015, updator=18614031015), WebsiteProps(id=1251199007909954106, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1251197148516266372, code=themeStyle, value=null, createTime=1776238689639, updateTime=1776238689639, creator=18614031015, updator=18614031015)])], journalTitle=重庆邮电大学学报(自然科学版), weixinUrl=null, journalUrl=http://journal2.cqupt.edu.cn/jcuptnsecn, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), journalPhotoCn=eIjswk9Qxcq7+V27dEZ90g==, journalPhotoEn=PN1NU8XEwLhVlBzwQEoroA==, 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/cqyddxxb/CN/10.3979/j.issn.1673-825X.202408040201, detailUrlEn=https://castjournals.cast.org.cn/joweb/cqyddxxb/EN/10.3979/j.issn.1673-825X.202408040201, pdfUrlCn=https://castjournals.cast.org.cn/joweb/cqyddxxb/CN/PDF/10.3979/j.issn.1673-825X.202408040201, pdfUrlEn=https://castjournals.cast.org.cn/joweb/cqyddxxb/EN/PDF/10.3979/j.issn.1673-825X.202408040201, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
融合跨时间共性特征的时序知识图谱推理模型
收藏切换
PDF下载
陈美琪 , 张诚麟 , 于洪
重庆邮电大学学报(自然科学版) | 人工智能与大数据 2025,37(5): 708-716
收起
收藏切换
重庆邮电大学学报(自然科学版) | 人工智能与大数据 2025, 37(5): 708-716
融合跨时间共性特征的时序知识图谱推理模型
全屏
陈美琪 , 张诚麟 , 于洪
作者信息
  • 重庆邮电大学 计算智能重庆市重点实验室,重庆 400065
  • 陈美琪,硕士研究生,主要研究方向为知识图谱和知识表示。E-mail:

    张诚麟,博士研究生,主要研究方向为因果学习和知识表示。E-mail:

    于洪,教授,博士,主要研究方向为粗糙集、粒计算、三支决策、智能信息处理、Web智能、数据挖掘。E-mail:

通讯作者:

Temporal knowledge graph reasoning model incorporating cross-time commonality features
Meiqi CHEN , Chenglin ZHANG , Hong YU
Affiliations
  • Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, P R China
doi: 10.3979/j.issn.1673-825X.202408040201
文章导航
收藏切换

时序知识图谱推理用于预测未出现在时序知识图谱中的事件信息,在推荐系统、问答系统和医疗等场景下取得了重要的应用。时序知识图谱中实体相关背景知识的缺失,限制了推理技术的发展。现有的推理方法依赖于外部图谱,却很少关注图谱自身所隐含的背景信息。为了充分挖掘图谱隐含的背景信息,通过抽取跨时间共性的特征刻画实体背景,提出了一种融合了跨时间共性特征的时序知识图谱推理模型(temporal knowledge graph reasoning model incorporating cross-time commonality features,TR-CTC)。TR-CTC利用图神经网络模型,从多跳路径中抽取跨时间共性特征作为实体的背景信息;融入到时序知识图谱表示学习过程中,提升推理性能。实验结果表明,TRCTC在链接预测任务中普遍优于基准模型。

时序知识图谱  /  图神经网络  /  链接预测  /  跨时间共性  /  知识图谱推理

Temporal knowledge graph reasoning, which predicts events absent from the graph, has seen significant applications in recommendation systems, question answering, and healthcare. The lack of background knowledge in temporal knowledge graphs hinders reasoning, with existing methods relying on external graphs while overlooking implicit data within the graph. To fully exploit the graph's implicit background information, this paper extracts cross-temporal features to define entity backgrounds and proposes a temporal knowledge graph reasoning model incorporating cross-time commonality features(TR-CTC). TR-CTC uses a graph neural network to extract cross-temporal commonality from multi-hop paths, integrating it as background information into the graph representation learning process, enhancing reasoning performance. Experimental results show that TR-CTC generally outperforms baseline models in link prediction tasks.

temporal knowledge graph  /  graph neural network  /  link prediction  /  cross-temporal commonality  /  knowledge graph reasoning
陈美琪, 张诚麟, 于洪. 融合跨时间共性特征的时序知识图谱推理模型. 重庆邮电大学学报(自然科学版), 2025 , 37 (5) : 708 -716 . DOI: 10.3979/j.issn.1673-825X.202408040201
Meiqi CHEN, Chenglin ZHANG, Hong YU. Temporal knowledge graph reasoning model incorporating cross-time commonality features[J]. Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition), 2025 , 37 (5) : 708 -716 . DOI: 10.3979/j.issn.1673-825X.202408040201
知识图谱在推荐系统[1]、问答系统[2]和医疗[3]等场景发挥了关键作用。典型的知识图谱[4-5]以三元组(sro)形式来表示人类知识,so对应于现实生活中的事物,r表示二元关系。知识图谱推理是发现知识图谱中潜在知识的过程[6-7]。随着实际应用场景中时间紧耦合数据的出现,时序知识图谱推理技术应运而生。时序知识图谱将三元组(sro)扩展为四元组(srot),其中t表示静态事件(sro)的时间有效性。
时序知识图谱推理分为补全和预测两个任务[8]。面向时序知识图谱补全任务的模型选择学习实体的浅嵌入,并考虑时间信息。例如,TTransE[9]、DisMult[10]、SimplE[11]、ChronoR[12]、Hyte[13]等模型通过将四元组信息映射为向量空间的变化信息完成编码。然而,仅对时序知识图谱进行补全推理已不能满足应用场景的需要[14]。最近,研究者们开始关注用于时序知识图谱预测任务的表示学习方法[15-23]。RE-Net[15]利用RGCN[16]对单个时间片的结构信息进行编码,在循环神经网络框架[17]上模拟时序的动态更新;EvoKG[18]对循环事件进行建模来刻画事件的演化过程;TeMP[19]、TiRGN[20]使用图卷积神经网对事件的时序规律性进行建模。这些方法基于时序知识图谱在时间维度上的信息描述实体的动态变化过程。除了时间维度上的信息,实体的背景信息也成为了研究者的另一种选择,它反映了图谱中实体在高阶上的关联。在时序知识图谱推理中,丰富实体的背景信息是为了更全面地表征实体及其动态特性,从而提高推理任务的准确性。现有方法从外部语义和自身结构两方面补充背景信息。RE-GCN[21]通过引入静态图约束组件来提供实体的背景信息。HiSMatch[22]利用CompGCN[23]对实体的背景知识图谱进行编码,进而编码实体的状态。上述方法因为在输入端需要额外引入外部语义补充背景信息,所以会受外部数据源的质量、一致性等条件的限制。时序知识图谱推理方法对时间片内邻接节点特征进行聚合来获取表示实体的动态属性,这类结构信息局限在单个时间片内。本文结合不受时间约束的结构,提出了将跨时间共性用于图谱推理的方法。
时序知识图谱本身隐含着实体的背景信息,其中包括跨时间共性。跨时间共性是指多个实体在特定情况下具有不受时间约束的静态属性。在时序知识图谱的不同时间段内,实体对之间存在完全相同的关系变化路径。ICEWS14数据集中的实例如图1所示。图1中,“Sultan bin Saeed Al Mansouri”与“阿拉伯联合酋长国”、“Wilben Mayor”与“菲律宾”、“Isa Ayo Salami”与“法院法官(尼日利亚)”之间都存在一条从“发表声明”到“逮捕、拘留或以法律行动起诉”的关系变化路径。这些路径上的事件分别发生在2014-06、2014-05和2014-01,属于不同的时间段。在时序知识图谱上,“Sultan bin Saeed Al Mansouri”;“Wilben Mayor”和“Isa Ayo Salami”等实体缺少详细的背景信息,现有方法只将他们看作独立存在的实体;但在现实世界中,他们分别联系各自的机关(如Wilben Mayor是菲律宾国家警署发言人),他们通过一系列的行动对各自的国家及下属机构产生影响。本文通过这样的关系变化路径,为两端的实体提取抽象的共性特征,把这些抽象的特征称为跨时间共性,能刻画出实体之间的高阶关联,提供实体缺失的背景信息。如Sultan bin Saeed Al Mansouri、Wilben Mayor和Isa Ayo Salami三个实体具有相似的背景信息。这类背景信息在该实例中可以被描述为“政府机关代表”。除了一条关系从“发表声明”变化到“逮捕、拘留或以法律行动起诉”的路径以外,三个实体还可能具有更多符合“政府机关代表”这种背景信息的关系变化路径。因此,本文将跨时间共性作为背景信息,以此丰富实体的信息。
本文提出了一种融合跨时间共性的时序知识图谱推理模型(temporal knowledge graph reasoning model incorporating cross-time commonality features,TR-CTC),包括时态路径采样、个性与共性特征融合和解码预测模块。利用图遍历生成时态路径,得到具有完全相同关系变化路径的实体对;在特征融合模块中,通过影响因子获取实体之间存在的跨时间共性,利用图卷积神经网络(graph convolutional network,GCN)融合单个时间片内的结构信息(个性特征)和跨时间共性(共性特征)。考虑到数据的时序性,本文用门控循环单元(gated recurrent unit,GRU)对特征进行更新,采用Conv-TransE[24]对特征进行解码预测,实现推理过程。
本文的主要贡献包括两个方面。
1)从新颖的视角出发,通过分析时序知识图谱中存在的多跳时态路径结构,挖掘跨时间共性作为实体的背景信息,为推理任务提供了丰富的信息支持。
2)充分利用图谱本身的信息,设计了基于图神经网络的融合模块,以GCN融合实体时间片内的结构特征和跨时间共性,进一步提升推理效果。
定义1 时序知识图谱。时序知识图谱t由四元组{(srot)|sort}组成,其中分别表示实体、关系和时间戳的集合。
定义2 时态路径。CTDNE[25]首次提出了时态路径的概念。时态路径中邻接的边需要满足时序递增的约束条件,保证采样路径中蕴含实体间动态交互的时序依赖信息。对于长度为n的时态路径[(sr1e1t1),(e1r2e2t2),…,(en-1rnentn)],路径中时间信息的偏序关系为t1t2≺…≺tn;关系变化路径为[r1r2,…,rn];实体变化路径为[se1e2,…,en-1en],其中se1,…,enr1r2,…,rn
TR-CTC采用编码器-解码器架构的框架如图2所示。TR-CTC包含三个组成单元,时态路径采样器、编码模块以及解码模块(预测模块)。
时态路径采样器通过图遍历的方式捕捉图中存在的多跳时态路径。编码模块的输入由两部分组成,分别为原始的时序知识图谱和从时序知识图谱中提取出来的时态路径。本文假设存在相同结构的实体之间具有跨时间的共性特征。在时序知识图谱中,TR-CTC将时态路径中相同关系变化路径当作具有相同结构的评判依据,以此获取存在跨时间共性的实体。TR-CTC利用影响因子计算这类实体之间存在的跨时间共性,利用GCN融合原始图谱中单个时间片内的结构信息(个性特征)和计算得到的跨时间共性(共性特征),生成更加丰富的实体特征和关系特征。同时,TR-CTC利用GRU进行时间维度上的更新。在解码阶段,TR-CTC采用ConvTransE[17]进行解码,完成预测任务。
时态路径采样旨在对多跳时态路径进行采样,在时间约束下对时序知识图谱进行遍历,获取满足时间偏序关系为t1t2≺…≺tnn跳时态路径[(sr1e1t1),(e1r2e2t2),…,(en-1rnentn)]。
特定实体的跨时间共性只与时态路径中具有相同关系变化路径的头、尾实体相关。对于实体s,在时间t采集到的时态路径集合Ps,t为在采样过程中,时态路径集合中的路径最后的四元组(en-1rnentn)满足时间tnt。其中,=[(sr1e1t1),(e1r2e2t2),…,(en-1rnentn)]。
路径起止对集合Cs,t为时态路径集合Ps,t中每一条路径的起始点和终止点集合,表示为
共性特征的提取与融合,是为了丰富实体的历史信息、强调实体潜在的状态。本文提出了一种新颖的GCN单元,在保留来自当前时间片实体个性特征的同时,融合了从多跳时态路径提取出来的跨时间共性特征。模块的输入为时间片内的时序知识图谱t,以及截止到该时间节点的所有路径Pt的路径起止对集合Ct,可得
式(3)—(4)中,{e1e2,…,en}是在Gt中的实体集合。与o具有相同关系变化路径的实体对表示为
式(5)中,ri=i=1,…,n
与候选节点相关的多跳路径反映了该实体的跨时间共性。考虑到每个实体对于候选实体的影响程度不同,本文设计了一种对跨时间共性进行编码的方式,即
式(6)中:^∈ℝ d在时间t实体o跨时间共性特征的第l层嵌入;No,t表示与o具有相同关系的路径起止对;表示节点ss′之间的相似程度,反映o′o的影响程度。表示为
式(7)中,d)是s在时间t的第l层嵌入∈ℝds′在时间t的第l层嵌入∈ℝd间的余弦相似度。本文称为影响因子。
为了融合实体的个性与共性特征,本文设计了一个聚合器。与RE-GCN[12]只考虑单跳邻居相比,本文不仅对单跳路径的结构信息进行编码,还能够将跨时间共性特征融入到实体特征的构造过程中。实体的特征聚合函数为
式(8)中:是实体so在时间t的第l层嵌入,为对应权重;Co是入度为o的归一化因子;σ是RReLu激活函数;φ是一维卷积算子。
为保留时序依赖关系,本文使用GRU在时间维度上对实体嵌入和关系嵌入进行更新。具体来说,利用当前时间片聚合得到的实体嵌入对前一个时间片的实体嵌入进行更新,面向实体嵌入的GRU定义为
式(9)中:HtHt-1tt-1处的d维实体嵌入组成的矩阵;是在t时刻单步卷积后的实体嵌入矩阵;GRU(·)为门控单元。
当前时间戳t的关系嵌入与当前时间片同r相关联的所有实体相关,将当前时间得到的关系嵌入设计为
式(10)中:rt是在t时连接到r的所有实体;rrR对应的嵌入;pooling(·)为平均池化操作,concat[;]为拼接操作。
为了保持与子图序列中实体嵌入更新的一致性,面向关系的GRU也将进行更新,表示为
式(11)中,由所有关系的组成,并且Rt-1Rt是在t-1和t时刻的关系嵌入矩阵。RtRt-1通过面向关系的GRU更新得到。
对于子图的特征嵌入,本文使用ConvTransE[17]模型来进行解码,评分函数为
式(12)—(13)中:posrt)、prsot)分别是预测实体和关系出现的概率函数;Hs,tHo,t分别是soHt中对应的嵌入;Rr,trRt中对应的嵌入;sigmoid(·)为激活函数。
模型需要对链接任务中的实体预测任务和关系预测任务进行训练。根据上述得到的概率预测函数分别计算两个任务的最小化交叉熵损失,即实体预测的损失Le和关系预测的损失Lr,其表达式为
式(14)—(15)中:T是训练集中的时间戳数量;分别表示两个任务的标签。
将两个任务看作是在多任务学习框架下进行的,因此总损失L
式(16)中,λ是控制损失占比的参数。
实验在集成危机预警系统(integrated crisis early warning system,ICEWS)[26]系列数据集上进行。该系列数据集是时序知识图谱链接预测的常用基准数据集。选择ICEWS14、ICEWS18和ICEWS0515子集作为实验数据集。数据集统计情况如表1所示。
实验旨在评估模型在时序知识图谱链接预测任务中的推理能力。具体来说,给定一个时序知识图谱t,链接预测任务中实体预测是在时刻t为关系r和实体e查询到一个实体e0,使得(ere0t)存在于时序知识图谱t;关系预测为实体e和实体e0查询关系r,使得(ere0t)在时序知识图谱t中存在。实验采用Hits@k和平均倒数排名(mean reciprocal rank,MRR)RMRR作为评价指标,这是基于知识图谱的链接预测任务的常用指标。两个指标的计算式为
式(17)—(18)中:Stest表示测试集中的四元组;rrank,i表示真实目标在预测结果中的排名;Ιrrank,ik)是指示函数,其值为1表示条件为真,否则为0。
本文使用TTransE[9]、DistMult[10]、SimplE[11]、RE-Net[15]、RE-GCN[21]和EvoKG[18]作为对比方法,验证提出模型的有效性。
在训练过程中,模型预测实体和关系的嵌入维度为200,对于ICEWS14、ICEWS18、ICEWS0515设置的最佳时间片长度分别为10、9、9。Adam用于参数学习,学习率设置为0.001,实体预测的损失占比为0.1~0.9。由于采样路径过程会导致图节点表示的过度平滑问题[27],因此将路径采样长度设置为3。
λ是实体预测任务损失和关系预测任务损失之间的可变权衡因素。为了探讨λ值的设定,本文进行了参数敏感性实验,结果如图3图4所示。当λ取值在0.7到0.9之间时,模型性能较好。在本实验中将λ设置为0.7。本文将H@k记为Hits@k的缩写。
表2表3分别给出了不同模型在ICEWS14、ICEWS18和ICEWS0515数据集上的实体预测和关系预测结果。
TR-CTC在三个数据集上的表现普遍优于基线模型,这样的结果验证了模型的有效性。具体来说,TR-CTC明显优于基于多维空间向量表示的TTransE、DisMult和SimplE模型。TR-CTC不仅捕捉了每个时间片上的实体和关系特征,还考虑了时间顺序带来的时序特征。因此,它可以捕获时序知识图谱上的演化模式。
TR-CTC在实体预测任务中的表现普遍优于外推预测推理模型RE-Net、RE-GCN和EvoKG。TRCTC通过捕捉多跳时态路径中存在的跨时间共性,所提取的实体特征表现力更强,邻域特征也更加丰富。在ICEWS0515数据集中,TR-CTC的MRR、H@1、H@ 3、H@ 10指标都有一定程度提升。对于ICEWS0515数据集,由于时间跨度比另外两个数据集大,因此能够提取跨时间共性的时态路径更加丰富,所提供的跨时间共性特征也就更加明显。对于ICEWS18数据集,由于其单个时间片的个性信息比另外两个数据集丰富,而EvoKG关注于对单个时间片的信息进行建模,这通常会导致其在ICEWS18数据集上有较高的H@1指标。
TR-CTC在关系预测任务中的总体性能普遍优于基线模型。对比发现,ICEWS14数据集中关系的数量小于ICEWS18和ICEWS0515数据集,所能提供的关系变化路径更多,效果提升更加明显。REGCN和EvoKG都采用关系聚合编码的方式,具有较高的H@10指标。
为进一步说明跨时间共性的必要性,本文进行了消融实验。实验结果如表4所示。实验中TRCTC表示本文提出的原始模型,w/o CTC表示去除跨时间共性之后的模型。在考虑了跨时间共性后,TR-CTC在实体预测和关系预测任务中表现出了更强的推理能力,并在三个数据集上都取得了提升,这与本文的假设一致。跨时间共性是对实体潜在的背景信息进行描述,所以融合实体背景信息的TRCTC模型在实体预测任务中性能提升更加明显。
为了验证基于时态路径挖掘实体背景信息方法的有效性,本文在实验中设计了TR-CTC(random)和TR-CTC(cos-sim)模式。TR-CTC(random)表示采用随机选择的实体替换路径的头实体和尾实体;TR-CTC(cos-sim)表示利用余弦相似度作为实体之间存在共性的衡量标准。在ICEWS14数据集上的实验结果如表5所示。从表5可以看到,TR-CTC(random)在实体预测和关系预测中MRR指标分别下降14.8百分点和1.7百分点;TR-CTC(cos-sim)的MRR指标分别下降4.14百分点和2.16百分点。TR-CTC在实体预测和关系预测中均优于TR-CTC(random)和TR-CTC(cos-sim)。该现象说明基于多跳时态关系变化路径提取的跨时间共性具有有效性。在实体预测中,跨时间共性的有效性表现得更明显,原因是跨时间共性作为实体的背景信息,丰富了实体的全局信息,辅助了推理任务的完成。
本文提出了一种融合跨时间共性的时序知识图谱推理方法,通过在多跳时态路径中捕捉跨时间共性来丰富实体表示。在ICEWS14、ICEWS18、ICEWS0515三个数据集上的实验结果证明了TRCTC在实体预测方面的显著优势和有效性,在关系预测方面有较为良好的性能。消融实验表明,跨时间共性在时序知识图谱推理中发挥着积极作用。未来的工作将尝试利用多种结构去挖掘时序知识图谱中潜在的关联信息,以提升时序知识图谱的推理性能。
参考文献 引证文献
排序方式:
[1]
ZHAO Y, WANG X, CHEN J, et al. Time-aware path reasoning on knowledge graph for recommendation[J]. ACM Transactions on Information Systems, 2022, 41(2): 1-26.
[2]
MAVROMATIS C, SUBRAMANYAM P L, IOANNIDIS V N, et al. Tempoqr: Temporal question reasonning over knowledge graphs[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Washington, USA: AAAI, 2022: 5825-5833.
[3]
陈德华,殷苏娜,乐嘉锦,.一种面向临床领域时序知识图谱的链接预测模型[J].计算机研究与发展, 2017, 54(12): 2687-2697.
CHENDH,YIN S N, LE J J, et al. A link prediction model for clinical temporal knowledge graph[J]. Journal of Computer Research and Development, 2017, 54(12):2687-2697.
[4]
BOLLACKER K, EVANS C, PARITOSH P, et al. Free base: A collaboratively created graph database for structuring human knowledge[C]//Proceedings of the 2008 ACM SIGMOD International Conference on Management of Data. New York, USA: ACM2008: 1247-1250.
[5]
FABIAN M, GJERGJI K, GERHARD W. Yago: A core of semantic knowledge unifying wordnet and wikipedia[C]//16th International World Wide Web Conference. Massachusetts, USA: WWW. 2007: 697-706.
[6]
SUN Z, DENG Z H, NIE J Y, et al. RotatE: Knowledge graph embedding by relational rotation in complex space[C]//International Conference on Learning Representations. Vancouver, Canada: ICLR, 2018: 1-5.
[7]
DETTMERS T, MINERVINI P, STENETORP P, et al. Convolutional 2d knowledge graph embeddings[C]//Proceedings of the AAAI Conference on Artificial Intelligence.Washington, USA: AAAI, 2018: 32-40.
[8]
王俞涵,陈子阳,赵翔,.时序知识图谱表示与推理的研究进展与趋势[J].软件学报, 2024, 35(8): 3923-3951.
WANG Y H, CHEN Z Y, ZHAO X, et al. Research progress and trend of temporal knowledge graph representation and reasoning[J]. Journal of Software, 2024, 35(8): 3923-3951.
[9]
JIANG T, LIU T, GE T, et al. Encoding temporal information for time-aware link prediction[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language processing. Austin, Texas, USA: EMNLP, 2016: 2350-2354.
[10]
GARCÍA-DURÁN A, DUMANCIC S, NIEPERT M. Learning sequence encoders for temporal knowledge graph completion[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. Brussels, Belgium: EMNLP, 2018: 4816-4821.
[11]
KAZEMI S M, POOLE D. SimplE embedding for link prediction in knowledge graphs[C]//Proceedings of the International Conference on Neural Information Processing Systems. Siem Reap, Cambodia: ICONIP, 2018: 4289-4300.
[12]
SADEGHIAN A, ARMANDPOUR M, COLAS A, et al. Chronor: Rotation based temporal knowledge graph embedding[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Washington, USA: AAAI, 2021:6471-6479.
[13]
DASGUPTA S S, RAY S N, TALUKDAR P. Hyte: Hyperplane-based temporally aware knowledge graph embedding[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural: Language Processing. Brussels, Belgium: EMNLP, 2018: 2001-2011.
[14]
MUTHIAH S, HUANG B, ARREDONDO J, et al. Planned protest modeling in news and social media[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Austin, Texas, USA: AAAI, 2015:3920-3927.
[15]
JIN W, QU M, JIN X, et al. Recurrent event network:Autoregressive structure inference over temporal knowledge graphs[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. Punta Cana, Dominican Republic: EMNLP, 2020: 6669-6683.
[16]
SCHLICHTKRULL M, KIPF T N, BLOEM P, et al. Modeling relational data with graph convolutional networks[C]//The Semantic Web: 15th International Conference. Heraklion Crete, Greece: Springer, 2018: 593-607.
[17]
CHO K, MERRIENBOER B V, GÜLÇEHREÇ, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. Doha, Qatar: EMNLP, 2014:1724-1734.
[18]
PARK N, LIU F, MEHTA P, et al. EvoKG: Jointly modeling event time and network structure for reasoning over temporal knowledge graphs[C]//Proceedings of the 15th ACM International Conference on Web Search and Data Mining. New York, USA: ACM, 2022: 794-803.
[19]
WU J, CAO M, CHEUNG J C K, et al. TeMP: Temporal message passing for temporal knowledge graph completion[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. Vienna, Austria: EMNLP, 2020: 5730-5746.
[20]
LI Y, SUN S, ZHAO J. TiRGN: Time-guided recurrent graph network with local-global historical patterns for temporal knowledge graph reasoning[C]//Proceedings of the 31th International Joint Conference on Artificial Intelligence. Vienna, Austria: IJCAI, 2022: 2152-2158.
[21]
LI Z, JIN X, LI W, et al. Temporal knowledge graph reasoning based on evolutional representation learning[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2021: 408-417.
[22]
LI Z, HOU Z, GUAN S, et al. HiSMatch: Historical structure matching based temporal knowledge graph reasoning[C]//Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. Abu Dhabi, United Arab Emirates: EMNLP, 2022: 7328-7338.
[23]
VASHISHTH S, SANYAL S, NITIN V, et al. Composition-based multi-relational graph convolutional networks[C]//International Conference on Learning Representations. Addis Ababa, Ethiopia: ICLR, 2020: 19-25.
[24]
SHANG C, TANG Y, HUANG J, et al. End-to-end structure-aware convolutional networks for knowledge base completion[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Honolulu, Hawaii, USA: AAAI, 2019: 3060-3067.
[25]
NGUYEN G H, LEE J B, ROSSI R A, et al. Continuoustime dynamic network embeddings[C]//Proceedings of the 2018 World Wide Web Conference. Lyon, France:ACM, 2018:969-976.
[27]
GIRALDO J H, SKIANIS K, BOUWMANS T, et al. On the trade-off between over-smoothing and over-squashing in deep graph neural networks[C]//Proceedings of the 32nd ACM International Conference on Information and knowledge management. Birmingham, United Kingdom:ACM, 2023: 566-576.
2025年第37卷第5期
PDF下载
168
85
引用本文
BibTeX
文章信息
doi: 10.3979/j.issn.1673-825X.202408040201
  • 接收时间:2024-08-04
  • 首发时间:2026-04-16
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-08-04
  • 修回日期:2025-09-06
基金
作者信息
    重庆邮电大学 计算智能重庆市重点实验室,重庆 400065

通讯作者:

参考文献
分享链接
https://castjournals.cast.org.cn/joweb/cqyddxxb/CN/10.3979/j.issn.1673-825X.202408040201
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