Research and Business Challenges in Data Mining Technology

Data mining aims at extracting hidden knowledge from the information stored in a data warehouse to identify the preferences of customers, derive decision models for business processes, or to describe the information stored in a condensed model. Techniques being researched fall into several categories, including search for association rules, clustering techniques, genetic algorithms, and classification algorithms. This paper provides an overview of a multi-strategy system, called Data Surveyor, to illustrate several research issues raised in developing data mining algorithms capable to exploit multi-gigabyte company databases using state-of-the-art DBMS technology. Furthermore, based on our experience in bringing the system to the market we identify the business challenges to implement and exploit this technology is real life settings.