摘要
Dataclusteringisasignificantinformationretrievaltechniqueintoday’sdataintensivesociety.Overthelastfewdecadesavastvarietyofhugenumberofdataclusteringalgorithmshavebeendesignedandimplementedforallmostalldatatypes.Thequalityofresultsofclusteranalysismainlydependsontheclusteringalgorithmusedintheanalysis.Architectureofaversatile,lessuserdependent,dynamicandscalabledataclusteringmachineispresented.Themachineselectsforanalysis,thebestavailabledataclusteringalgorithmonthebasisofthecredentialsofthedataandpreviouslyuseddomainknowledge.Thedomainknowledgeisupdatedoncompletionofeachsessionofdataanalysis.
出版日期
2009年02月12日(中国Betway体育网页登陆平台首次上网日期,不代表论文的发表时间)