TY - JOUR
T1 - Learning Graph Representation with Generative Adversarial Nets
AU - Wang, Hongwei
AU - Wang, Jialin
AU - Wang, Jia
AU - Zhao, Miao
AU - Zhang, Weinan
AU - Zhang, Fuzheng
AU - Li, Wenjie
AU - Xie, Xing
AU - Guo, Minyi
N1 - Publisher Copyright:
© 1989-2012 IEEE.
PY - 2021/8/1
Y1 - 2021/8/1
N2 - Graph representation learning aims to embed each vertex in a graph into a low-dimensional vector space. Existing graph representation learning methods can be classified into two categories: generative models that learn the underlying connectivity distribution in a graph, and discriminative models that predict the probability of edge between a pair of vertices. In this paper, we propose GraphGAN, an innovative graph representation learning framework unifying the above two classes of methods, in which the generative and the discriminative model play a game-Theoretical minimax game. Specifically, for a given vertex, the generative model tries to fit its underlying true connectivity distribution over all other vertices and produces 'fake' samples to fool the discriminative model, while the discriminative model tries to detect whether the sampled vertex is from ground truth or generated by the generative model. With the competition between these two models, both of them can alternately and iteratively boost their performance. Moreover, we propose a novel graph softmax as the implementation of the generative model to overcome the limitations of traditional softmax function, which can be proven satisfying desirable properties of normalization, graph structure awareness, and computational efficiency. Through extensive experiments on real-world datasets, we demonstrate that GraphGAN achieves substantial gains in a variety of applications, including graph reconstruction, link prediction, node classification, recommendation, and visualization, over state-of-The-Art baselines.
AB - Graph representation learning aims to embed each vertex in a graph into a low-dimensional vector space. Existing graph representation learning methods can be classified into two categories: generative models that learn the underlying connectivity distribution in a graph, and discriminative models that predict the probability of edge between a pair of vertices. In this paper, we propose GraphGAN, an innovative graph representation learning framework unifying the above two classes of methods, in which the generative and the discriminative model play a game-Theoretical minimax game. Specifically, for a given vertex, the generative model tries to fit its underlying true connectivity distribution over all other vertices and produces 'fake' samples to fool the discriminative model, while the discriminative model tries to detect whether the sampled vertex is from ground truth or generated by the generative model. With the competition between these two models, both of them can alternately and iteratively boost their performance. Moreover, we propose a novel graph softmax as the implementation of the generative model to overcome the limitations of traditional softmax function, which can be proven satisfying desirable properties of normalization, graph structure awareness, and computational efficiency. Through extensive experiments on real-world datasets, we demonstrate that GraphGAN achieves substantial gains in a variety of applications, including graph reconstruction, link prediction, node classification, recommendation, and visualization, over state-of-The-Art baselines.
KW - Graph representation learning
KW - generative adversarial nets
KW - graph softmax
UR - http://www.scopus.com/inward/record.url?scp=85077249788&partnerID=8YFLogxK
U2 - 10.1109/TKDE.2019.2961882
DO - 10.1109/TKDE.2019.2961882
M3 - Article
AN - SCOPUS:85077249788
SN - 1041-4347
VL - 33
SP - 3090
EP - 3103
JO - IEEE Transactions on Knowledge and Data Engineering
JF - IEEE Transactions on Knowledge and Data Engineering
IS - 8
M1 - 8941296
ER -