Analyzing HCI issues in data clustering tools

Abstract

Due to the rapid growth in the volume of data stored in organizational databases and the human limitations in analyzing and interpreting data, appropriate technics are necessary to allow the identification of a large amount of information and knowledge in such databases. In this context, several techniques and tools have been proposed for enabling the end user to interpret his dataset. In this work we discuss the ways of interacting with cluster analysis tools, taking into account both the clustering and the interpretation stages. We investigate how usability and user experience aspects of such tools can improve the understanding of the discovered knowledge. Moreover, we evaluate the role of visualization methods in the comprehension of groups formed in cluster analysis using Knime, Orange Canvas, RapidMiner Studio and Weka data mining tools. © 2014 Springer International Publishing.