I’d thought about how to present my second layer of Categories for a long time since I started to simplify my classification of sub-categories.
Here is the structure of my sub-categories based on the first layer.
Besides presenting the proportion of each sub-category, like this:
I also want to show the layers before and after the sub-categories where is spacially as close to it as possible. So I come up a structure like this:
Take the category “Figure & Figures” as an example.
The layer before its sub-categories is the changes in the proportion of “Figure & Figures” among all the categories in the first layer from 2003 to 2018.
And the layer after the description of the components of sub-categories is an array of charts that show how the proportion of each sub-category changed every four years from 2003 to 2018.
In order to squeeze this structure onto a 2d medium, I need to use some on-screen behaviors to connect these three layers — for this case, I think I just need “hover over” and “click”: hover over the pie chart to see the layer before, and make each legend as a button to toggle the charts in the layer after.
However, some categories in the first layer might have too many sub-categories and the proportion of some sub-categories is too small to show the changes by time. So I decided to just show four sub-categories of each category in the first layer which have the largest proportion(expert category “NA”).
Here is the simplified version of the structure of sub-categories:
For the number of records in the category “Personal life”, “Education” and “Sports” are seemingly less than the records in other first categories, it might make more sense to just narrate them with texts instead of classifying their sub-categories, most of which might only have one record throughout 16 years.
