Research map / Graph learning / Knowledge and temporal graphs
Relation Structure-Aware Heterogeneous Information Network Embedding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use relation-specific Euclidean modeling for affiliation relations and translation modeling for interaction relations | RHINE · DBLP, Yelp, AMiner node clustering | NMI: 0.6065 -> 0.7204; 0.3657 -> 0.3882; 0.5586 -> 0.6024 | Helped |
| Use relation-specific Euclidean modeling for affiliation relations and translation modeling for interaction relations | RHINE · DBLP, Yelp, AMiner link prediction | AUC: 0.9160 -> 0.9315; 0.8987 -> 0.9148; 0.8653 -> 0.8762; 0.9141 -> 0.9316; 0.8902 -> 0.9173 | Helped |
| Use relation-specific Euclidean modeling for affiliation relations and translation modeling for interaction relations | RHINE · DBLP, Yelp, AMiner multi-class classification | Macro-F1: 0.8976 -> 0.9344; 0.6836 -> 0.7132; 0.9962 -> 0.9884 | Mixed |
| Use relation-specific Euclidean modeling for affiliation relations and translation modeling for interaction relations | RHINE · DBLP, Yelp, AMiner multi-class classification | Micro-F1: 0.9000 -> 0.9250; 0.7399 -> 0.7572; 0.9965 -> 0.9807 | Mixed |
| Model every relation with Euclidean distance instead of differentiating relation structures (ablation) | RHINE · Node clustering, link prediction, and classification | task performance | Hurt |
| Model every relation with translation instead of differentiating relation structures (ablation) | RHINE · Node clustering, link prediction, and classification | task performance | Hurt |
| Reverse the relation-specific models: Euclidean for interaction relations and translation for affiliation relations (ablation) | RHINE · Node clustering, link prediction, and classification | task performance | Hurt |
| Use translation for all relations rather than Euclidean distance (ablation) | RHINE · Node clustering, link prediction, and classification | task performance versus all-Euclidean variant | Helped |
| Increase embedding dimension (ablation) | RHINE · Node clustering | NMI | Helped |
| Increase negative sample count (ablation) | RHINE · Node clustering | NMI | Helped |
Closest papers by meaning
- Hyperbolic Heterogeneous Information Network Embedding (AAAI 2019 · similarity 0.92)
- An Attention-Based Graph Neural Network for Heterogeneous Structural Learning (AAAI 2020 · similarity 0.90)
- Heterogeneous Graph Neural Network on Semantic Tree (AAAI 2025 · similarity 0.88)
- HeterSEED: Semantics–Structure Decoupling for Heterogeneous Graph Learning under Heterophily (NeurIPS 2026 · similarity 0.86)
- Deep Structural Knowledge Exploitation and Synergy for Estimating Node Importance Value on Heterogeneous Information Networks (AAAI 2024 · similarity 0.85)
- Link Prediction in Multilayer Networks via Cross-Network Embedding (AAAI 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.