A data mining design framework - A preview

Abstract

Data storages contain a lot of hidden information unknown to their owners. There are many different types of data mining processes, which aim to provide the means to expose this hidden information. But existing data mining processes mainly illustrate the management process and not really the discovery process. The user still has to decide, which methods and algorithms to apply. Furthermore, correct interpretation of the result can be challenging. In this paper we describe a framework for a systematic knowledge discovery process, which is split into three stages. We define the requirements, methods, and possible outcomes for each stage. © 2010 Springer-Verlag.