Network Embedding-Based Approach for Detecting Collusive Spamming Groups on E-Commerce Platforms

Algorithm 1

The modified DeepWalk algorithm.

	Input: HIN , window size , embedding size , the number of walks per node , walk length , set of user nodes
	Output: matrix of user node representations
(1)	Initialization: Sample from
(2)	Build a binary tree from
(3)	for i = 0 to n do
(4)	← Shuffle ()
(5)	for each do
(6)	← RandomWalk (, , l)
(7)	←
(8)	for each node do
(9)	ifthen
(10)	add to
(11)	end if
(12)	end for
(13)	SkipGram (, , )
(14)	end for
(15)	end for
(16)	return