Data Mining and Knowledge Discovery In Databases

Exploring approaches and tools for data cleaning in data mining and knowledge discovery

Abstract

There are many approaches for data cleaning. Some of themare: parsing3, datatransformation, duplicateelimination and statisticalmethod. A large variety of tools is available in the market to supportdata transformation and data cleaning tasks, in particular for datawarehousing. Some tools concentrate on a specific domain, such as cleaning nameand address data, or a specific cleaning phase, such as data analysis orduplicate elimination. Due to their restricted domain, specialized toolstypically perform very well but must be complemented by other tools to addressthe broad spectrum of transformation and cleaning problems.