User’s Attitude under the Perspective of Mental Energy Flow

Algorithm 1

ELMFluid (G, p, K).

	Input: The rating network G; The parameters for the initialization p; K time slots
	Output:The predicted ratings of the target non-raters;
	Initialize:
(1)	Associate each node in G with a container and each edge with a directed pipe;
(2)	Add an extra pipe that can inject fluid in all containers;
(3)	Initialize the fluid height and temperature in each container;
(1)	for i = 0 to K do
(2)	for each non-rater’s container do
(3)	if >then
(4)	Update the volume and temperature according to equations (6) and (7);
(5)	end if
(6)	end for
(7)	for each rater’s container do
(8)	Inject additional fluid to maintain the fluid height and temperature;
(9)	end for
(10)	end for
(11)	return the fluid temperature in the target container.