Research map / Probabilistic and scientific modeling / Dynamical systems
Inductive Representation Learning in Temporal Networks via Causal Anonymous Walks
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 CAW-N instead of the strongest baseline | Temporal link prediction · New vs. new links on Reddit, Wikipedia, MOOC, Social Evolution, Enron, and UCI | AUC: CAW-N: 97.71 ± 0.08; 99.29 ± 0.10; 90.86 ± 0.56; 95.55 ± 0.40; 96.98 ± 0.41; 93.16 ± 0.25. Strongest baselines: 94.96 ± | Helped |
| Add set-based CAW node anonymization and encoding f1(I_CAW) (ablation) | CAW-N-mean · Wikipedia, UCI, and Social Evolution inductive all test links | AUC: Wikipedia: 96.28 ± 0.66 -> 98.94 ± 0.11; UCI: 79.45 ± 0.71 -> 91.88 ± 0.26; Social Evolution: 53.69 ± 0.21 -> 91.97± 0.2 | Helped |
| Add temporal encoding f2(t) (ablation) | CAW-N-mean · Wikipedia, UCI, and Social Evolution inductive all test links | AUC: Wikipedia: 97.90 ± 0.17 -> 98.94 ± 0.11; UCI: 83.91 ± 0.39 -> 91.88 ± 0.26; Social Evolution: 68.25 ± 1.99 -> 91.97± 0.2 | Helped |
| Replace CAW set-based identities with one-hot anonymous-walk identities (ablation) | CAW-N-mean · Wikipedia, UCI, and Social Evolution inductive all test links | AUC: Wikipedia: 98.94 ± 0.11 -> 96.55 ± 0.21; UCI: 91.88 ± 0.26 -> 85.59 ± 0.34; Social Evolution: 91.97± 0.23 -> 68.47 ± 0.8 | Hurt |
| Set alpha=0 for uniform history sampling (ablation) | CAW-N-mean · Wikipedia, UCI, and Social Evolution inductive all test links | AUC: Wikipedia: 98.94 ± 0.11 -> 74.38 ± 2.05; UCI: 91.88 ± 0.26 -> 80.44 ± 0.56; Social Evolution: 91.97± 0.23 -> 77.72 ± 0.4 | Hurt |
Closest papers by meaning
- Neural Temporal Walks: Motif-Aware Representation Learning on Continuous-Time Dynamic Graphs (NeurIPS 2022 · similarity 0.90)
- Improving Temporal Link Prediction via Temporal Walk Matrix Projection (NeurIPS 2024 · similarity 0.86)
- Continuous-Time Graph Representation with Sequential Survival Process (AAAI 2024 · similarity 0.86)
- Learning the Causal Structure of Networked Dynamical Systems under Latent Nodes and Structured Noise (AAAI 2024 · similarity 0.85)
- MOCHA: Discovering Multi-Order Dynamic Causal Structure in Temporal Point Processes (NeurIPS 2026 · similarity 0.85)
- Causal Temporal Representation Learning with Nonstationary Sparse Transition (NeurIPS 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.