IGT: Modeling local and global patterns of graph data by interleaving GNNs and transformers

Keyphrases

• Graph Neural Network 100%
• Global Patterns 100%
• Neural Transformer 100%
• Graph Data 100%
• Local Pattern 100%
• Graph Transformer 75%
• Transformer 50%
• Message Passing 50%
• Local Operator 50%
• Information Integration 25%
• Competitive Performance 25%
• Graph Representation 25%
• Text Classification 25%
• Transformer Model 25%
• Graph Learning 25%
• Learning Task 25%
• Transformer Layer 25%
• Transformer Framework 25%
• Transformer Module 25%
• Aspect-based Sentiment Analysis 25%
• Unified Graph 25%
• Graph Tasks 25%
• Graph Representation Learning 25%

Computer Science

• Graph Neural Network 100%
• Graph Transformer 75%
• Message Passing 50%
• Local Operator 50%
• Relative Performance 25%
• Global Information 25%
• Text Classification 25%
• Graph Representation 25%
• Transformer Model 25%
• Local Information 25%
• Graph Representation Learning 25%
• Aspect-Based Sentiment Analysis 25%