|
Author
Bagué i Collell, Maria
|
Abstract
After conducting a literature search on the concept of visualization taxonomy, no established or generalized classification criteria have been identified, but each author presents his proposal with his own criteria.
The objective of this study is the elaboration of a taxonomy of visualizations based on the studied bibliography and, also, the realization of a first validation attempt. To this end, this work aims to establish a fairly general taxonomy that serves as a starting point regardless of the scope of application.
After having studied the taxonomies identified as most important in this field and trying to detect if any relevant reference was missing by building a network of citations, a summary table of the bibliography has been prepared that allows to extract the main characteristics of each taxonomy studied. This has been the first input to begin to elaborate the own taxonomic proposal that has four levels: Objective to communicate, Number of variables, Type of variable and Volume of data. Each of the visualizations that are part of the taxonomy with the R program have been exemplified. The proposal has been subjected to a validation process with two different methods: by comparison with the visualizations that appear in some reference magazines and from the opinions of the experts.
Once the results of the two validations have been obtained, it is concluded that the taxonomic proposal is valid with coincidences greater than 50%, although it has some exceptions. The last and fourth level Volume of data is questioned in the validation by experts since the branch of the first level Distribution only has 31.58% of coincidences and is where the level Volume of data is sensitive. This level could be studied in future research, where it should be checked whether to eliminate or replace it with another, such as the number of categories to be displayed in the case of categorical variables.
|
|
