GraphGAN: Graph Representation Learning with Generative Adversarial Nets

By Hongwei Wang, Jia Wang, Jialin Wang, Miao Zhao, Weinan Zhang, Fuzheng Zhang, Xing Xie, Minyi Guo. 2017

698 citations recorded by the source corpus.

The goal of graph representation learning is 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 the graph, and discriminative models that predict the probability of edge existence between a pair of vertices. In this paper, we propose GraphGAN, an innovative graph representation learning framework unifying above two classes of methods, in which the generative model and 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, when considering the implementation of generative model, we propose a novel graph softmax 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 link prediction, node classification, and recommendation, over state-of-the-art baselines.

Subfields: Artificial Intelligence, Molecular Biology, Statistical and Nonlinear Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Molecular Biology

In Statistical and Nonlinear Physics

Citations in the corpus

RippleNet: Propagating User Preferences on the Knowledge Graph for Recommender Systems (2018, methodology): “addition, our work also incorporates a KGE term as a regularizer for more stable and effective learning. 3.6.3 Knowledge Graph Embedding. RippleNet also connects to a large body of work in KGE methods[3, 10, 14, 19, 28, 30, 37, 41]. KGE intends to embed entities and relations in a KG into continuous”
DKN: Deep Knowledge-Aware Network for News Recommendation (2018, methodology): “In addition to translation-based methods [4, 18, 26, 48] used in DKN, researchers have also proposed many other models such as Structured Embedding [5], Latent Factor Model [17], Neural Tensor Network [37] and GraphGAN [42].”
A Comprehensive Survey on Graph Neural Networks (2019, neutral): “llenge/ 12707 - - - 12 [23], [56] Unstructured Graphs MNIST yann.lecun.com /exdb/mnist/ 70000 - - - 10 [12], [21], [24], [26], [55] Wikipedia www.mattmahoney .net/dc/textdata 1 4777 184812 - 40 [65], [114] 20NEWS [120] 1 18846 - - 20 [12], [42] Others METR-LA [121] - - - - - [29], [73] Movie-Lens1M ”
NetGAN: Generating Graphs via Random Walks (2018, methodology): “The recently proposed GraphGAN (Wang et al., 2017) is another instance of such prescribed edge-level probabilistic models, where f is optimized using the GAN objective instead of the traditional cross-entropy.”
Learning Graph Embedding With Adversarial Training Methods (2019, background): “However, their approach can only exploit the topological information [47], [49], [50].”
Embedding Temporal Network via Neighborhood Formation (2018, background): “Besides, the strengths of generative adversarial networks [8] have also been exploited with network embedding [6, 26].”
Sample-Efficient Optimization in the Latent Space of Deep Generative Models via Weighted Retraining (2020, background): “…be chosen to be a state-of-the-art deep generative model (DGM), such as a variational autoencoder (VAE) [28, 46] or a generative adversarial network (GAN) [14], which have been shown to be capable of learning vector representations of many types of high-dimensional, structured data [52, 2, 59, 7].”
Amalgamating Knowledge from Heterogeneous Graph Neural Networks (2021, background): “Graph neural networks (GNNs) have achieved unprecedented advances in recent years, showing promising performance in handling graph data lying in the non-Euclidean domain [16, 37, 53, 44, 6, 41, 19, 23, 36, 13, 18, 25, 42].”
Conditional Structure Generation through Graph Variational Generative Adversarial Nets (2019, background): “Most existing graph generation models like [18, 12, 7, 40, 25, 52] are only generating links among fixed set of nodes, and thus do not satisfy Eq.”
Deep Learning on Graphs: A Survey (2018, neutral): “[105] Regularization for GAEs O ð M Þ No GCPN [116] Rewards for Graph RL O ð MN Þ Yes MolGAN [117] Rewards for Graph RL O ð N 2 Þ Yes GraphGAN [124] Generation of negative samples (i.e., node pairs) O ð MN Þ No ANE [125] Regularization for network embedding O ð N Þ No GraphSGAN [126] Enhancing…”
Network Representation Learning: A Survey (2017, methodology): “For each sampled path starting from vi and ending with vj, the representations are learned by maximizing the target vertex vj’s occurrence probability conditioned on the source vertex vi PrðvjjviÞ ¼ expðrsvi rtvjÞP v2V expðrsvi rtvÞ : (11) GraphGAN.”
Self-supervised Heterogeneous Graph Neural Network with Co-contrastive Learning (2021, background): “HeCo_GAN GAN-based models [10, 14, 34] play a minimax game between a generator and a discriminator, and aim at forcing generator to generate fake samples, which can finally fool a welltrained discriminator.”
A Survey on Heterogeneous Graph Embedding: Methods, Techniques, Applications and Sources (2020, neutral): “In homogeneous graph, the adversarial-based techniques only consider the structural information, for example, GraphGAN [127] uses Breadth First Search when generating virtual nodes.”
Semi-supervised Learning on Graphs with Generative Adversarial Nets (2018, background): “In GraphGAN [36], generator learns embeddings for nodes and discriminator solves”