|< Day Day Up >|| |
This study is intentionally exploratory and the objective is to discover a process for performing data mining projects. The case study suggests that the observed data mining activities are not completely off the mark with the straw man of what a process for data mining projects might be. The value of this study is that it contributes to a practical and applied understanding of data mining and proposes a process that may be used as a starting point when making decisions about planning, organizing, executing and closing data mining projects. In the absence of such a starting point, overall project risk may increase and project planning and execution are not optimized. It is a well-known fact in the product development arena that a product must first be commercialized before revenue can be generated and for that matter not all commercialized products guarantee a positive revenue stream. The existence of a disciplined process and quality execution of activities within the entire process have been found to positively impact NPD projects. The NPD literature is thus the basis for a proposed four-phase end-to-end process for performing data mining projects. This four-phase process, which begins with a Discovery phase and ends with an Execution and Infusion phase, makes no assumptions about an organization's prior experience with data mining technology and may equally apply to both in-house and outsourced data mining projects. The 23 activities in the proposed process can serve as a starter set of activities for consideration and can be modified depending on the nuances of the particular project under question.
Advances in database technology, computer processors and the push towards electronic commerce have meant that our ability to generate, collect and store data has outstripped our ability to discern new and valuable information from existing data. The infamous "store now, extract later" thinking continues to reign in many organizations. The natural question many senior managers and executives are beginning to ask is: how can I exploit this vast sea of data to improve my relations with my customer and to maintain a competitive edge over my competitors? This same question will undoubtedly take on greater importance over the next few years. Consequently, data mining will become a major corporate imperative in leading organizations.
The objective of this study is to discover a process for performing data mining projects and to propose this process as a starting point. Since the nature of this study was exploratory, no claims of comprehension can be made. An obvious initial direction for future research is to undertake a cross-sectional case study analysis to corroborate and build on this study. Once the proposed process is evaluated and possibly enhanced, a quantitative study, using a survey instrument, can be undertaken to test the effectiveness of such a process. Prescriptions can then be made about "how" to perform data mining, which is a necessary first step in a research agenda concerned with developing a comprehensive data mining methodology.
Second, the role of data mining in the context of a larger knowledge management process needs to be explored. Knowledge management linked with corporate strategy will become more important as the rate of technological change increases, customer loyalty diminishes and lead times for developing new and improved products and services is reduced. Data mining as a form of technological innovation may play an important role in driving a knowledge management strategy that is integrated with an organization's corporate strategy. Finally, globalization and advances in information technology mean that the reliance on using colocated teams will diminish and the subsequent reliance on virtual teams will deepen. Virtual teams are groups of geographically and/or organizationally dispersed collaborators that are brought together for a specific purpose and whose members' primary mode of interaction is through communications-related information technology. The use of virtual teams in data mining implementation projects has not been studied in the past and is certainly a relevant avenue of research to pursue.
|< Day Day Up >|| |