According to the paper《node2vec: scalable feature learning for networks》.
The current random walk algorithm requires 2 additional features.
Biased random walk.
Second order random walk.
Biased random walk, control the probability of walking edges through edge weight.
Second order random walk, add 2 parameters to control the walking strategy.
(1) parameter p, called return factor. Controls the probability of re-walk to a previously walked vertex
(2) parameter q, called in-out factor. Controls whether to walk inward or outward.
Feature Description (功能描述)
According to the paper《node2vec: scalable feature learning for networks》. The current random walk algorithm requires 2 additional features.
Biased random walk, control the probability of walking edges through edge weight. Second order random walk, add 2 parameters to control the walking strategy. (1) parameter p, called return factor. Controls the probability of re-walk to a previously walked vertex (2) parameter q, called in-out factor. Controls whether to walk inward or outward.
Please refer to the paper design diagram:
P1
P2