Abstract:Node embedding techniques have gained prominence since they produce continuous and low-dimensional features, which are effective for various tasks. Most existing approaches learn node embeddings by exploring the structure of networks and are mainly focused on static non-attributed graphs. However, many real-world applications, such as stock markets and public review websites, involve bipartite graphs with dynamic and attributed edges, called attributed interaction graphs. Different from conventional graph data, attributed interaction graphs involve two kinds of entities (e.g. investors/stocks and users/businesses) and edges of temporal interactions with attributes (e.g. transactions and reviews). In this paper, we study the problem of node embedding in attributed interaction graphs. Learning embeddings in interaction graphs is highly challenging due to the dynamics and heterogeneous attributes of edges. Different from conventional static graphs, in attributed interaction graphs, each edge can have totally different meanings when the interaction is at different times or associated with different attributes. We propose a deep node embedding method called IGE (Interaction Graph Embedding). IGE is composed of three neural networks: an encoding network is proposed to transform attributes into a fixed-length vector to deal with the heterogeneity of attributes; then encoded attribute vectors interact with nodes multiplicatively in two coupled prediction networks that investigate the temporal dependency by treating incident edges of a node as the analogy of a sentence in word embedding methods. The encoding network can be specifically designed for different datasets as long as it is differentiable, in which case it can be trained together with prediction networks by back-propagation. We evaluate our proposed method and various comparing methods on four real-world datasets. The experimental results prove the effectiveness of the learned embeddings by IGE on both node clustering and classification tasks.

Learning Temporal Interaction Graph Embedding Via Coupled Memory Networks.

TigeCMN: On exploration of temporal interaction graph embedding via Coupled Memory Neural Networks

Learning Node Embeddings in Interaction Graphs

IGE+: A Framework for Learning Node Embeddings in Interaction Graphs.

Temporal Interaction Embedding for Link Prediction in Global News Event Graph

TI-GCN: A Dynamic Network Embedding Method with Time Interval Information

Learning dynamic embeddings for temporal attributed networks

TME-BNA: Temporal Motif-Preserving Network Embedding with Bicomponent Neighbor Aggregation

Continuous-Time Dynamic Graph Learning Via Neural Interaction Processes.

TempNodeEmb:Temporal Node Embedding considering temporal edge influence matrix

Learning Multiple Temporal Relational Network Embeddings Via Graph Convolutional Network.

TIGER: Temporal Interaction Graph Embedding with Restarts.

Dynamic Graph Representation Learning Via Coupling-Process Model.

Multi-Relation Aware Temporal Interaction Network Embedding

Discrete-time Temporal Network Embedding via Implicit Hierarchical Learning in Hyperbolic Space

Temporal Graph Transformer for Dynamic Network

Exploring Temporal Information for Dynamic Network Embedding

GTEA: Inductive Representation Learning on Temporal Interaction Graphs via Temporal Edge Aggregation

ConTIG: Continuous representation learning on temporal interaction graphs

Graph-Level Embedding for Time-Evolving Graphs

Embedding Models for Episodic Knowledge Graphs