Migrating Datain Telecom Billing
Himanshu Agrawal, Girish B. Chafle, Sunil Goyal, Sougata Mukherjea · 2008
Datamigration hasbecomea priority inmany industries, spawnedbyavariety ofbusiness needs. Mostofthe existing tools forExtract, Transform andLoad(ETL)process ofdatamigration arepiece-meal anddonotpresent acomplete solution. Moreover, whileresearch hasfocused ontheproblem ofSchemaMapping, a keystepintheETL process, most ofthecurrent algorithms donotperform wellonreal-world data. Researchers havesuggested theuseofDomainKnowledge toenhanceschemamapping. Inthispaper, we usedomain knowledge inaninnovative mannertoimprove schemamapping inan'actual' industrial setting. Further, wetakeacomprehensive viewofthedatamigration problem andpresent anend-to-end system fortheETL process, utilizing existing tools foreachstep andbuilding connectors, wherever required. We focus onData Migration forTelecomBilling andutilize domainknowledge captured inan ontology, a thesaurus anda setofrulesto improve schemamapping. Experiments conducted onareal-life datademonstrate theeffectiveness ofoursystem andvalidate the utility ofdomainknowledge indatamigration projects.