Database Marketing Process Supported by Ontologies: An Oil Company Distribution Network Case Study

Knowledge Discovery in Databases

Knowledge discovery in databases (KDD) is the result of an exploratory process in order to achieve domain defined objectives involving the application of various algorithmic procedures for manipulating data, building models from data, and manipulating the models. The Data Mining phase deserves more attention from the research community: processes comprise multiple algorithmic components, which interact in nontrivial ways.

We consider tools that will help data analysts to navigate the space of KDD processes systematically, and more effectively. In particular, this paper focuses on a subset of stages of the KDD —those stages for which there are multiple algorithm components that can apply.

For most of this paper, we consider a prototypical KDD process template, similar to the one represented in Figure 1.The sequence of KDD phases is not strict. Moving back and forth between different phases is always required. It depends on the outcome of each phase, which one, or which particular task of a phase has to be performed next.

Figure 1.

Knowledge discovery process framework. Adapted from (Fayyad et al., 1996).

We focus our attention on the three main macro components of KDD life cycle: data understanding (data selection); data pre processing (all related data preparation and transformation activities), and modeling (data mining and the application of induction algorithms) We have chosen this set of components because, individually, they are relatively well understood—and they can be applied to a wide variety of benchmark data sets.