yuwvandy / dpgnn Goto Github PK

Hello Dr. Wang, I am re-reading your paper intensively. I have encountered some problems in the explanation part of formula (9). I would like to ask you to solve confusion.

1.“computes the inverse ratio of the number of labeled nodes in the node 𝑣𝑖’s class 𝜙(𝑣𝑖),Í𝑣𝑗∈V𝑙Y𝑗𝜙(𝑣𝑖), to the total number of labeled nodes |V𝑙|.”My colleagues have different perceptions of this sentence, We want to know what exactly Í𝑣𝑗∈V𝑙Y𝑗𝜙(𝑣𝑖) stands for?

2.why |V𝑙| and Í𝑣𝑗∈V𝑙Y𝑗𝜙(𝑣𝑖) are inversely ratio, Shouldn't they be proportional when the weighting factors are given?

3.Then this part also talks about the network homogeneity can expand the training data, I would like to ask you whether it is achieved by formula 12?

Sorry to bother you, looking forward to your reply!!

In the process of running the code, I found that the DE, EN, ES....
I don’t know which data sets these names replace.
Could you please answer in your busy schedule?