EGAT: Evolving Graph Attention Network Weights for Dynamic Link Prediction

5.5 KB of data related to the EGAT (Evolving Graph Attention Networks) model, which evolves the weight of k attention heads at each time step for dynamic link prediction. The dataset was authored by Yucai Jiang and last updated on June 1, 2026. It is shared under a CC-BY-4.0 license on figshare.

Use Cases

• Benchmarking dynamic graph neural network architectures based on the described evolving attention head mechanism.
• Analyzing the temporal evolution of attention weights in graph networks for link prediction.
• Training or validating models for predicting future connections in time-varying networks.

Strengths

• Dataset is small (5.5 KB), facilitating quick download and inspection.
• Has a clear, permissive license (CC-BY-4.0) for reuse.
• Associated with a specific, named model (EGAT) for dynamic link prediction.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Yucai Jiang via figshare

Freshness

Last updated 2026-06-01 17:40:52; freshness should be verified.