Translating Advances in Data Mining to Business Operations: The Art of Data Mining in Retailing

Plummeting data storage costs mean that businesses can now hold more data than ever before. However, increased volume and detail of data necessitates effective and efficient analytical tools. Knowledge Discovery in Databases (KDD) is a field of research that studies the development and use of various data analysis tools and techniques. Data mining is one such tool. KDD research has produced an array of models, theories, functions and methodologies for producing knowledge from data. However, despite these advances, nearly two thirds of IT managers say that data mining products are too difficult to use in a business context. This paper discusses how advances in data mining translate into the business context. It highlights the art of business implementation rather than the science of KDD. INTRODUCTION In the past, high storage and processing costs meant that businesses had to be selective about what data they stored. Today, this restriction has been removed as costs of data storage plummet – the money that bought 40 megabytes of storage in 1989 buys 8 gigabytes today [2]. The arrival of holographic storage will enable further reductions in the cost of storage capacity [3] and fuel the already popular data warehousing phenomenon. As the volume and detail of stored data increases, the demand for effective and efficient analysis tools also increases [4]. KDD has been rigorously researched, particularly in the area of data mining [5]. This has resulted in an array of models, theories, functions and methodologies for producing knowledge from data. However, despite these advances, nearly two thirds of IT managers say that data mining products are too difficult to use in a business context [1]. Aside from the scientific aspects of KDD, lies the artistic application of KDD to business. Limited academic attention has been paid to the business implementation of specific data mining techniques [6]. This paper focuses on the broad issue of how advances in data mining translate into business use. We first present a review of the relevant academic literature and then apply this knowledge to a business case study in a major petrol (gas) service station corporation LITERATURE REVIEW The purpose of this literature review is to provide an understanding of the concepts of data mining to be used in discussing the case study.