Title:
|
HEATMAP MATRIX: USING REORDERING, DISCRETIZATION AND FILTERING RESOURCES TO ASSIST MULTIDIMENSIONAL DATA ANALYSIS |
Author(s):
|
Miguel Mechi Naves Rocha and Celmar Guimarães da Silva |
ISBN:
|
978-989-8704-42-9 |
Editors:
|
Yingcai Xiao, Ajith Abraham, Guo Chao Peng and Jörg Roth |
Year:
|
2022 |
Edition:
|
Single |
Keywords:
|
Multidimensional Data Visualization, Matrix Visualization, Reorderable Matrix, Data Mining, Association Rules, Discretization |
Type:
|
Full Paper |
First Page:
|
11 |
Last Page:
|
18 |
Language:
|
English |
Cover:
|
|
Full Contents:
|
click to dowload
|
Paper Abstract:
|
Visualization can help users to analyze multidimensional datasets and make better decisions related to them. The current literature presents a set of multidimensional data visualization techniques that deal with discrete variables, continuous ones, or both. Recently, the heatmap matrix was proposed as an alternative for visualizing multidimensional datasets with discrete variables. It comprises a matrix of heatmaps, in which each heatmap shows the frequency of the values of two variables in the records of a dataset. Despite its proposed utility, it only presents a static representation of the dataset, which may limit users understanding about the dataset under analysis. Therefore, we propose to add interactive resources to this visualization, aiming to provide distinct views of a dataset. To guide the choice of new features, we compared the resources of a set of multidimensional visualizations and selected the characteristics that are suitable to enhance the heatmap matrix. We added to it resources for: (a) reordering values (rows and columns) of variables in its heatmaps; (b) filtering variables and values based on correlations of variables and on association rules; (c) discretizing continuous variables to be represented in the visual structure; among others. We present two case studies in which we show the potential of the heatmap matrix, enhanced by the resources proposed in this work, to assist the analysis of multidimensional data. |
|
|
|
|