资源描述：
《基于mapreduce的关联规则数据挖掘在电信网络运营分析中的应用》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。
1、基于MapReduce的关联规则数据挖掘在电信网络运营分析中的应用????????????????//.paper.edu.cn基于MapReduce的关联规则数据挖掘在电信网络运营分析中的应用苗宇，侯春萍北京邮电大学大学电子工程学院，北京市 100876摘要：本文基于电信运营商建设大数据共享平台的应用需求和现状，研究相关技术，提出了利用开源框架Hadoop对海量信令数据进行挖掘，以弥补传统作业系统在这方面的缺陷与不足，设计并实现了基于MapReduce的并行化Apriori算法，通过关联规则挖
2、掘发现移动通信网络业务量和业务质量的数据突变状况。实验结果表明，该方法能够有效地从海量数据中发现数据突变的相关业务因素和场景，为上层应用有针对性解决数据突变问题提供依据。关键词：数据挖掘;关联规则;海量数据;Hadoop;运营分析中图分类号：TP391ApplicationofMapReduce-BasedAssociationRuleDataMiningonOperationofTelecommunicationNetworkMiaoYu,HouChun-PingDepartmentofEle
3、ctronicEngineering,BeijingUniversityofPostsandTelecommunications,Beijing100876Abstract:Thispaperbasedontherequirementsandstatusofconstructingdata-sharingplatformfortelecomoperators,proposedusingtheHadoopopensourceframeworkformassivesignalingdataminin
4、ganalysistomakeupforthedefectsanddeficienciesoftraditionalmonitoringsystem.AHadoop-basedsignalingminingplatformandaMapReduce-basedAprioriAlgorithmisdesignedandimplementedtodiscovertheassociationrulesonnetworkservicequalityanddatamutation.Theexperimen
5、talresultsshowthat,themethodcaneffectivelydiscoveravarietyofbusinessfactorsandscenarioswhichleadtothemutation,andprovidesupportfortheupperapplicationtosolvetheproblem.Keywords:DataMining;AssociationRule;BigData;Hadoop;OperationAnalysis-1-????????????
6、????//.paper.edu.cn0IntroductionWiththerapiddevelopmentofmobilecommunication,coexistingof2G/3G/4Gsystem,appearenceofcomlpextypesoftelecommunicationequipments,anddispersivemodeofcon-straction,thetraditionalnetworkmanagementsystemofoperatorsarefacinggr
7、eatchallenges.Networkmanagementsystemoughttobecloud-based,integrated,intelligent,automaticinfuture.OneofthesignificantdirectionsfortheevolutionofNMSis3-layeredarchitectureincludedNetworkElementandEquipmentLayer,AcquisitionlayerandData-SharingLayer.Am
8、ongthem,theData-SharingLayerisresponsibleforintegratingvarioustypesofdataandprovidingunifieddataserviceforallofupperlayerapplications.Thereforeoperatorswillhaveanaturaladvantageofcollectingbigdata[1].However,asthedramaticincreasingindatasize,thetradi
