setting alpha and beta

[armintabari]

For the experiments in the paper you noted that you set alpha as 1, and beta as degree(i)^-1. I am trying to change and use the code here (as I understand both alpha and beta are assumed to be 1 here), but what exactly do you mean by degree(i)^-1? Is it the angle between the word being updated and the current neighbor? I'd appreciate it if you could clear this up for me.

[armintabari]

Could you please clarify this? Thank you.

[nooralahzadeh]

@armintabari
For each neighbour of i , beta is equal to degree(i)^-1 means that beta=1/number of neighbours degree(i) =number of neighbours since alpha=1, when you put in formula it becomes: new_vector_for_word= [(1/number of neighbours)*(SUM_all_neighbours_vector) + pre_vector_for_word ] / 2 or [SUM_all_neighbours_vector + number of neighbours*pre_vector_for_word]/(2*number of neighbours)

[armintabari]

Thank you.