Visual techniques for the interpretation of data mining outcomes

Abstract

The visual senses for humans have a unique status, offering a very broadband channel for information flow. Visual approaches to analysis and mining attempt to take advantage of our abilities to perceive pattern and structure in visual form and to make sense of, or interpret, what we see. Visual Data Mining techniques have proven to be of high value in exploratory data analysis and they also have a high potential for mining large databases. In this work, we try to investigate and expand the area of visual data mining by proposing a new 3-Dimensional visual data mining technique for the representation and mining of classification outcomes and association rules. © Springer-Verlag Berlin Heidelberg 2005.