Abstract:Modeling heterogeneity by extraction and exploitation of high-order information from heterogeneous information networks (HINs) has been attracting immense research attention in recent times. Such heterogeneous network embedding (HNE) methods effectively harness the heterogeneity of small-scale HINs. However, in the real world, the size of HINs grow exponentially with the continuous introduction of new nodes and different types of links, making it a billion-scale network. Learning node embeddings on such HINs creates a performance bottleneck for existing HNE methods that are commonly centralized, i.e., complete data and the model are both on a single machine. To address large-scale HNE tasks with strong efficiency and effectiveness guarantee, we present Decentralized Embedding Framework for Heterogeneous Information Network (DeHIN) in this paper. In DeHIN, we generate a distributed parallel pipeline that utilizes hypergraphs in order to infuse parallelization into the HNE task. DeHIN presents a context preserving partition mechanism that innovatively formulates a large HIN as a hypergraph, whose hyperedges connect semantically similar nodes. Our framework then adopts a decentralized strategy to efficiently partition HINs by adopting a tree-like pipeline. Then, each resulting subnetwork is assigned to a distributed worker, which employs the deep information maximization theorem to locally learn node embeddings from the partition it receives. We further devise a novel embedding alignment scheme to precisely project independently learned node embeddings from all subnetworks onto a common vector space, thus allowing for downstream tasks like link prediction and node classification. As shown from our experimental results, DeHIN significantly improves the efficiency and accuracy of existing HNE models as well as outperforms the large-scale graph embedding frameworks by efficiently scaling up to large-scale HINs.

Key Nodes Cluster Augmented Embedding for Heterogeneous Information Networks.

Self-supervised Nodes-Hyperedges Embedding for Heterogeneous Information Network Learning

Embedding Heterogeneous Information Network in Hyperbolic Spaces

WMPEClus: Clustering Via Weighted Meta-Path Embedding for Heterogeneous Information Networks.

M-HIN: Complex Embeddings for Heterogeneous Information Networks via Metagraphs

Proximity-aware Heterogeneous Information Network Embedding

Homogenization with Explicit Semantics Preservation for Heterogeneous Information Network

Relation Structure-Aware Heterogeneous Information Network Embedding

HAHE: Hierarchical Attentive Heterogeneous Information Network Embedding.

Heterogeneous Information Network Hashing for Fast Nearest Neighbor Search

Dynamic Heterogeneous Information Network Embedding With Meta-Path Based Proximity

Node and Edge Joint Embedding for Heterogeneous Information Network

Enhancing Network Embedding with Implicit Clustering

Heterogeneous Information Network Embedding with Convolutional Graph Attention Networks.

Meta-HGT: Metapath-aware HyperGraph Transformer for heterogeneous information network embedding

(HTNE)-T-2: Temporal Heterogeneous Information Network Embedding in Hyperbolic Spaces

H<SUP>2</SUP>TNE: Temporal Heterogeneous Information Network Embedding in Hyperbolic Spaces

A Simple Meta-path-free Framework for Heterogeneous Network Embedding

DeHIN: A Decentralized Framework for Embedding Large-Scale Heterogeneous Information Networks

Estimating Node Importance Values in Heterogeneous Information Networks

Prototype-Enhanced Hypergraph Learning for Heterogeneous Information Networks