Data pre-processing for database marketing

Abstract

To increase effectiveness in their marketing and Customer Relationship Manager activities, many organizations are adopting strategies of Database Marketing (DBM). Nowadays, DBM faces new challenges in business knowledge since current strategies are mainly approached by classical statistical inference, which may fail when complex, multi-dimensional and incomplete data is available. An alternative is to use Knowledge Discovery from Databases (KDD), which aims at automatic extraction of useful patterns by using Data Mining (DM) techniques. When applied to DBM, the identified patterns can be used for the efficient characterization of the customers. This paper focus several problems that arose in the data pre-processing step (e.g. data cleaning), which is necessary for the success of the DM approach to a DBM project.