图数据管理与图机器学习：协同效应与机会 Graph Data Management and Graph Machine Learning - Synergies and Opportunities.docx

下载文档

0
0
约9.27万字
约 28页
2025-02-08 发布于山西
举报
版权申诉
保障服务

图数据管理与图机器学习：协同效应与机会 Graph Data Management and Graph Machine Learning - Synergies and Opportunities.docx

1、本文档共28页，可阅读全部内容。
2、有哪些信誉好的足球投注网站（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。
3、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。

GraphDataManagementandGraphMachineLearning:SynergiesandOpportunities

ArijitKhan1XiangyuKe2YinghuiWu3

1AalborgUniversity,Denmark2ZhejiangUniversity,China3CaseWesternReserveUniversity,USA

arXiv:2502.00529v1[cs.DB]1Feb20251arijitk@cs.aau.dk2xiangyu.ke@3yxw1650@

arXiv:2502.00529v1[cs.DB]1Feb2025

ABSTRACT

Theubiquityofmachinelearning,particularlydeeplearn-ing,appliedtographsisevidentinapplicationsrang-ingfromcheminformatics(drugdiscovery)andbioin-formatics(proteininteractionprediction)toknowledgegraph-basedqueryanswering,frauddetection,andso-cialnetworkanalysis.Concurrently,graphdataman-agementdealswiththeresearchanddevelopmentofef-fective,ef?cient,scalable,robust,anduser-friendlysys-temsandalgorithmsforstoring,processing,andanalyz-ingvastquantitiesofheterogeneousandcomplexgraphdata.Oursurveyprovidesacomprehensiveoverviewofthesynergiesbetweengraphdatamanagementandgraphmachinelearning,illustratinghowtheyintertwineandmutuallyreinforceeachotheracrosstheentirespec-trumofthegraphdatascienceandmachinelearningpipeline.Speci?cally,thesurveyhighlightstwocru-cialaspects:(1)Howgraphdatamanagementenhancesgraphmachinelearning,includingcontributionssuchasimprovedgraphneuralnetworkperformancethroughgraphdatacleaning,scalablegraphembedding,ef?cientgraph-basedvectordatamanagement,robustgraphneu-ralnetworks,user-friendlyexplainabilitymethods;and(2)howgraphmachinelearning,inturn,aidsingraphdatamanagement,withafocusonapplicationslikequeryansweringoverknowledgegraphsandvariousdatasci-encetasks.Wediscusspertinentopenproblemsandde-

lineatecrucialresearchdirections.

1.INTRODUCTION

Graphdata,rangingfromsocialandbiologicalnet-worksto?nancialtransactions,knowledge