A Conceptual Model for Guiding the Clustering Analysis

Abstract

Abstract. Knowledge discovery from databases, in the descriptive approach, includes clustering analysis (CA) as an alternative to estimate how a set of objects is organized in the space of their dimensions. The main objective in this task is to find "natural" groups that could exhibit some meaning. Considering the strong subjectivity that underlies this process, an important issue refers to the relationships among the CA players when looking for a model that could adjust the data. In this work, a model for actions coordination that provides an order to drive the relationships among CA players is presented. This model is presented as a conceptual contribution towards the construction of a computational environment to support effective conversations in a subjective context.