An Open-World Extension to Knowledge Graph Completion Models
This work addresses the challenge of open-world link prediction for knowledge graph applications, though it is incremental as it extends existing embedding-based methods.
The authors tackled the problem of enabling knowledge graph completion models to predict facts for entities not seen during training by using textual descriptions, achieving competitive results on datasets like FB20k and DBPedia50k.
We present a novel extension to embedding-based knowledge graph completion models which enables them to perform open-world link prediction, i.e. to predict facts for entities unseen in training based on their textual description. Our model combines a regular link prediction model learned from a knowledge graph with word embeddings learned from a textual corpus. After training both independently, we learn a transformation to map the embeddings of an entity's name and description to the graph-based embedding space. In experiments on several datasets including FB20k, DBPedia50k and our new dataset FB15k-237-OWE, we demonstrate competitive results. Particularly, our approach exploits the full knowledge graph structure even when textual descriptions are scarce, does not require a joint training on graph and text, and can be applied to any embedding-based link prediction model, such as TransE, ComplEx and DistMult.