摘要
Currently,schemaintegrationframeworksuseapproacheslikerule-based,machinelearning,etc.Thispaperpresentsanontology-basedwrapper-mediatorframeworkthatusesboththerule-basedandmachinelearningstrategiesatthesametime.Theproposedframeworkusesglobalandlocalontologiesforresolvingsyntacticandsemanticheterogeneity,andXMLforinteroperability.Theconceptsinthecandidateschemasaremergedonthebasisofthesimilaritycoefficient,whichiscalculatedusingthedefinedrulesandthepriormappingsstoredinthecase-base.
出版日期
2005年06月16日(中国期刊网平台首次上网日期,不代表论文的发表时间)