基于开放网络知识的信息检索与数据挖掘.pdfVIP

计算 机研究 与发展 DOI：10．7544／issn1OOO一1239．2015 JournalofComputerResearchandDevelopment 52(2)：456—474，2015 基于开放网络知识的信息检索与数据挖掘 王元卓 贾岩涛 刘大伟。靳小龙 程学旗 (中国科学院网络数据科学与技术重点实验室 (中国科学院计算技术研究所) 北京 100190) z(烟 台中科 网络技术研究所 ／中国科学 院计算技术研究所烟台分所 山东烟台 264005) (wangyuanzhu0@ ict．ac．cn) OpenW ebKnowledgeAidedInformationSearchandDataM ining WangYuanzhuo ，JiaYantao，LiuDawei。，JinXiaolong ，andChengXueqi (CASKeyLaboratoryofNetworkDataScience Technology (InstituteofComputingTechnology，ChineseAcademyof Sciences)，Beijing100190) (InstituteofNetworkTechnology／ICT(YANTAI)，ChineseAcademyofSciences，Yantai，Shandong264005) Abstract Network bigdatareferstothemassivedatageneratedviainteraction and fusion ofthe ternaryhuman—machine—thinguniversein thecyberspaceandavailableontheInternet．Ithasafew typicalfeatures，such asmulti—sourced，heterogeneous，interactive，bursty，and noisy． Itcontains mainly unstructured data， and has strong real—timeness． Network big data implicitly contains tremendoushighly—interconnectedknowledge．Buildingup openW eborientedlarge—scaleknowledge basesisaneffectivemeansforobtainingrichknowledgefrom networkbigdata．Thispapercompares both thedomesticandinternationalmainstream openW ebknowledgebases．W especificallyanalyze thecoretechniquesandmethodsforconstructingopenW eb knowledgebases，fusingmulti～sourced knowledge，andupdatingtheknowledgebases．Furthermore，wesummarizetheresearch statusand main issues of open W eb knowledge base based information search， data mining， and system applicationsfrom differentaspects，includinguserintensionunderstanding，queryextension，semantic QA ，cluemining，relationshipreferencing，andpredictionofrelationshipsandattributes．Finally， welookintothedevelopmenttrendsandmainchallengesofopenW ebknowledgebases． Keywords networkbigdata；openW eb knowledge；ontology；informationsearch；datamining 摘 要 网络大数据是指 “人 、机 、物”三元世界在 网络空间