A Relational Tucker Decomposition for Multi-Relational Link Prediction
This work addresses link prediction in knowledge graphs, offering a more flexible framework that could improve performance, though it appears incremental as it builds on prior decomposition methods.
The authors tackled multi-relational link prediction in knowledge graphs by proposing the Relational Tucker3 decomposition, which generalizes existing models and allows flexible parameterization. Their experiments showed that automatically learned sparsity patterns and dense models can outperform sparse ones in some datasets.
We propose the Relational Tucker3 (RT) decomposition for multi-relational link prediction in knowledge graphs. We show that many existing knowledge graph embedding models are special cases of the RT decomposition with certain predefined sparsity patterns in its components. In contrast to these prior models, RT decouples the sizes of entity and relation embeddings, allows parameter sharing across relations, and does not make use of a predefined sparsity pattern. We use the RT decomposition as a tool to explore whether it is possible and beneficial to automatically learn sparsity patterns, and whether dense models can outperform sparse models (using the same number of parameters). Our experiments indicate that---depending on the dataset--both questions can be answered affirmatively.