Transactions on Machine Learning and Data MiningP-ISSN: 1865-6781, E-ISSN 2509-9337ISBN 978-3-942952-78-1
Volume 13 - Number 2 - October 2020 - Pages 83-99
A Visualization Framework for Post-Processing of Association Rule Mining
Manal A. Alyobi and Arwa A. JamjoomKing Abdulaziz University Jeddah, Saudi Arabia
Abstract
Association rule mining is one of the most used techniques in data mining. It can be applied in different fields to uncover the hidden relationships in large data sets. Despite that fact, association rule mining algorithms often leave the analyst with the task of analyzing and understanding thousands of rules. Also, this large number of rules cannot be represented effectively in classical visualizations. Thus, the current interest in visualizing data mining is toward applying visual data exploration techniques. These techniques enable the user to interact with graphs. In this paper we introduce a new visualization technique for association mining results based on tree methods, specifically the collapsible tree. To do that, we converted a grouped matrix into a collapsible tree by utilizing R packages. We demonstrate our technique through a market basket analysis case study.
Keywords: Visual Data Mining, Association Rule Mining, Collapsible Tree, Market Basket Analysis.
Download Paper (813.2 KB)