Understanding ROC curve in decision tree

[#1]

I use standard Car Poll.jmp dataset to build decision tree to predict 'Country' by 'Size'. I split decision tree twice and then I display ROC plt. I clicked on question mark and then on ROC plot and read explanation very carefully. Unfortunately it is still unclear to me how ROC is constructed. Does anybody know any good resources explaining construction of ROC plot? I do have general understanding of ROC curve, but in case of decision tree, I am confused.

[#2 · Accepted solution]

Here is the information from .........Help==>Books==>Predictive and Specialized Modeling

12871_pastedImage_0.png