Criteria for a Comparative Study of Visualization Techniques in Data Mining.pdf

Proceedings of the IEEE 3rd International Conference On Intelligent Systems Design and Application, Tulsa, USA,(ISDA 2003) ISBN: 3540404260 Springer-Verlag, Berlin Criteria for a Comparative Study of Visualization Techniques in Data Mining Robert Redpath and Bala Srinivasan School of Computer Science and Software Engineering,Faculty of Information Technology,Monash University,Caulfield East, Victoria, Australia Abstract.The paper considers the relationship of information visualization tools to the data mining process. The types of structures that may be of interest to data mining experts in data sets are outlined. The performance of a particular visualiza- tion technique in revealing those structures and supporting the subsequent steps in the process needs to be based on a number of criteria. The criteria for performing an evaluation are suggested and explained. A division into two main groups is suggested; criteria that relate to interface issues and criteria that relate to the char- acteristics of the data set. An example application of some of the criteria is made. Keywords.Information visualization, data mining, comparative criteria, evalua- tion, multi-dimensional, patterns, visualization software 1 Introduction Information visualization can be a useful first step in the data mining process. It can act as an exploratory tool to guide the choice of some other algorithmic tech- nique or it can be used to isolate some subset of the larger data set for more de- tailed analysis. It may also directly reveal structures of interest in the data set un- der consideration. In order to judge the effectiveness of any particular visualization technique criteria need to be established upon which the success or failure of each technique can be measured. This paper extends the criteria on which visualization techniques can be compared and categorizes