|
Path Breaking Case Studies in E-commerce using Data MiningKeywords: Web analytics , retail e-commerce , Simpson’s paradox , Timeout Analysis , bot analysis. Abstract: — The e-commerce domain can provide all the rightingredients for successful data mining and claim that it is a killerdomain for data mining. The architecture of various e-commercesites has supported data collection, transformation, and datamining since its inception. With click-streams being collected atthe application-server layer, high-level events being logged, anddata automatically transformed into a data warehouse usingmeta-data, common problems plaguing data mining usingweblogs (e.g., sessionization and conflating multi-sourced data)are obviated, thus allowing one to concentrate on actual datamining goals. The paper briefly reviews the architecture ofintegrated E-Commerce with Data Mining, discusses some casestudies and puts forward some conclusions inferred from thesame. While the conclusions are drawn from the case studiesfrom the retail e-commerce domain, they are also equallyapplicable to other data mining domains, as well.
|