An Improved Sequential Recommendation Algorithm based on Short-Sequence Enhancement and Temporal Self-Attention Mechanism

Algorithm 1

The sequential recommendation algorithm proposed in this paper.

	Input: The behavior sequence of user
	Output: The recommendation list result of user , denoted as
(1)	for in do
(2)	ifthen
(3)
(4)	else
(5)	%Date preprocessing
(6)	end if
(7)	end for
(8)	for in do
(9)	%Reverse prediction model training
(10)	end for
(11)	for in do
(12)	%Short sequence enhancement
(13)	end for
(14)	for in do
(15)	Generate time interval matrix;
(16)	Calculate time-aware self-attention model;
(17)	Apply the point-wise feed-forward network and further processed;
(18)	Calculate prediction and loss;
(19)	end for
(20)	return ;