FIsViz: A frequent itemset visualizer

Abstract

Since its introduction, frequent itemset mining has been the subject of numerous studies. However, most of them return frequent itemsets in the form of textual lists. The common cliché that "a picture is worth a thousand words" advocates that visual representation can enhance user understanding of the inherent relations in a collection of objects such as frequent itemsets. Many visualization systems have been developed to visualize raw data or mining results. However, most of these systems were not designed for visualizing frequent itemsets. In this paper, we propose a frequent itemset visualizer (FIsViz). FIsViz provides many useful features so that users can effectively see and obtain implicit, previously unknown, and potentially useful information that is embedded in data of various real-life applications. © 2008 Springer-Verlag Berlin Heidelberg.