Title: Visualization of cross tabulation by the association rules by using the correspondence analysis
Authors: Yoshiro Yamamoto - Tokai Univeristy (Japan) [presenting]
Sanetoshi Yamada - Tokai University (Japan)
Abstract: When comparing the response in the survey by gender and age, we make the cross-tabulation tables then visualize them by such like mosaic plot. For many answers to multiple-choice item, we want to find the item that the reaction of a particular layer (gender and age) is different from the others. Association rule analysis are suitable for this kind of analysis. By using the coordinates by correspondence analysis it is possible to plot the relationship between items and media layers. In this visualization, correspondence analysis and association rule analysis are complementary to each other. In addition, by showing the percentage of respondents each item and each layer, it becomes possible to understand the trend between items and layers. The visualization is constructed by using RStudio Shiny. It is possible to change the various parameters of the association rule analysis interactively.