Heterogeneous Deep Graph Infomax

Ren, Yuxiang; Liu, Bo; Huang, Chao; Dai, Peng; Bo, Liefeng; Zhang, Jiawei

doi:10.48550/arxiv.1911.08538

Cited by 27 publications

(29 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We found out that competing methods treat the same dataset in various different ways to report their performance. In order to show that our method is efficient under different configurations, we employ two different evaluation protocols (EP) as proposed in [12] (EP1) and in [2] (EP2).…”

Section: Experimental Methodologymentioning

confidence: 99%

“…This strategy can be applied to any homogeneous self-supervised method to handle the heterogeneity of the graph; for example, we later extend GIC to heterogeneous GIC (HGIC). HDGI [12] acts in similar manner to obtain compact representations, which are optimized with the DGI objective. DMGI [9] enhances the DGI objective with a consensus regularization so that meta-path representations are similar to each other.…”

Section: Related Workmentioning

confidence: 99%

“…For competing methods, we report their performance by reusing the reported results of [12,2]. Whenever we need to implement a method by our own, we do it based on their original implementations and proposed configurations 2 , and for a fair comparison, we use the same encoders for these methods as explained below.…”

Section: Reproducibilitymentioning

confidence: 99%

See 2 more Smart Citations

HeMI: Multi-view Embedding in Heterogeneous Graphs

Mavromatis,

Karypis

2021

Preprint

View full text Add to dashboard Cite

Many real-world graphs involve different types of nodes and relations between nodes, being heterogeneous by nature. The representation learning of heterogeneous graphs (HGs) embeds the rich structure and semantics of such graphs into a low-dimensional space and facilitates various data mining tasks, such as node classification, node clustering, and link prediction. In this paper, we propose a self-supervised method that learns HG representations by relying on knowledge exchange and discovery among different HG structural semantics (meta-paths). Specifically, by maximizing the mutual information of meta-path representations, we promote meta-path information fusion and consensus, and ensure that globally shared semantics are encoded. By extensive experiments on node classification, node clustering, and link prediction tasks, we show that the proposed self-supervision both outperforms and improves competing methods by 1% and up to 10% for all tasks.Heterogeneous graph [17]. A heterogeneous graph (HG), denoted as G = (V, E), consists of an object set V and a link set E. It is used to represent a

show abstract

Section: Experimental Methodologymentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

HeMI: Multi-view Embedding in Heterogeneous Graphs

Mavromatis,

Karypis

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Then an averaging operation is adopted to calculate the global graph representation ℎ. Although there exist several techniques to generate the graph summary vector ℎ, the simple averaging operation demonstrates superior performance [32], and thus ℎ is calculated as…”

Section: Embeddingmentioning

confidence: 99%

Detecting Communities from Heterogeneous Graphs: A Context Path-based Graph Neural Network Model

Luo,

Fang,

Cao

et al. 2021

Preprint

View full text Add to dashboard Cite

Community detection, aiming to group the graph nodes into clusters with dense inner-connection, is a fundamental graph mining task. Recently, it has been studied on the heterogeneous graph, which contains multiple types of nodes and edges, posing great challenges for modeling the high-order relationship between nodes. With the surge of graph embedding mechanism, it has also been adopted to community detection. A remarkable group of works use the meta-path to capture the high-order relationship between nodes and embed them into nodes' embedding to facilitate community detection. However, defining meaningful meta-paths requires much domain knowledge, which largely limits their applications, especially on schema-rich heterogeneous graphs like knowledge graphs. To alleviate this issue, in this paper, we propose to exploit the context path to capture the high-order relationship between nodes, and build a Context Path-based Graph Neural Network (CP-GNN) model. It recursively embeds the high-order relationship between nodes into the node embedding with attention mechanisms to discriminate the importance of different relationships. By maximizing the expectation of the co-occurrence of nodes connected by context paths, the model can learn the nodes' embeddings that both well preserve the high-order relationship between nodes and are helpful for community detection. Extensive experimental results on four real-world datasets show that CP-GNN outperforms the state-of-the-art community detection methods 1 . CCS CONCEPTS• Information systems → Data mining.

show abstract

“…HeGAN [21] employs a generative adversarial approach for HIN embedding. Following DGI [52], HDGI [37] adopts a contrastive approach based on mutual information maximization. NSHE [60] is another approach to learn HIN embedding based on multiple meta-paths.…”

Section: Related Work 21 Unsupervised Network Embeddingmentioning

confidence: 99%

A Framework for Joint Unsupervised Learning of Cluster-Aware Embedding for Heterogeneous Networks

Khan,

Kleinsteuber

2021

Preprint

View full text Add to dashboard Cite

HIN) embedding refers to the low-dimensional projections of the HIN nodes that preserve the HIN structure and semantics. HIN embedding has emerged as a promising research field for network analysis as it enables downstream tasks such as clustering and node classification. In this work, we propose VaCA-HINE for joint learning of cluster embeddings as well as cluster-aware HIN embedding. We assume that the connected nodes are highly likely to fall in the same cluster, and adopt a variational approach to preserve the information in the pairwise relations in a cluster-aware manner. In addition, we deploy contrastive modules to simultaneously utilize the information in multiple meta-paths, thereby alleviating the meta-path selection problem -a challenge faced by many of the famous HIN embedding approaches. The HIN embedding, thus learned, not only improves the clustering performance but also preserves pairwise proximity as well as the high-order HIN structure. We show the effectiveness of our approach by comparing it with many competitive baselines on three real-world datasets on clustering and downstream node classification. CCS CONCEPTS• Information systems → Clustering and classification; Clustering and classification; • Theory of computation → Unsupervised learning and clustering; • Mathematics of computing → Variational methods; • Computing methodologies → Learning latent representations.

show abstract

Heterogeneous Deep Graph Infomax

Cited by 27 publications

References 23 publications

HeMI: Multi-view Embedding in Heterogeneous Graphs

HeMI: Multi-view Embedding in Heterogeneous Graphs

Detecting Communities from Heterogeneous Graphs: A Context Path-based Graph Neural Network Model

A Framework for Joint Unsupervised Learning of Cluster-Aware Embedding for Heterogeneous Networks

Contact Info

Product

Resources

About