A new conceptual model for dynamic text clustering Using unstructured text as a case

Abstract

In recent years, clustering has become a critical success factor for data analysis. Most clustering methods are sensitive to outliers, noise, presentation order, configuration architecture, Bellman's curse of dimensionality and complex shapes. They use the cost functions to reflect the general knowledge about internal structures and distributions of target data. There is no provided mechanism to reflect the dynamics of clustering environment on the data set. Hence, in the present study, an alternative numerical scheme (SC) was proposed to enhance the predictive accuracy of clustering. Our approach exploits variables selection techniques and Fuzzy Adaptive Resonance Theory to increase productivity of knowledge extraction.