Unified Interpretation of Smoothing Methods for Negative Sampling Loss Functions in Knowledge Graph Embedding

Open in new window