What Makes Entities Similar? A Similarity Flooding Perspective for Multi-sourced Knowledge Graph Embeddings
This work addresses the lack of interpretability in entity alignment for knowledge graphs, which is incremental as it bridges existing embedding models with a conventional algorithm.
The paper tackles the problem of understanding how entity alignment works in multi-sourced knowledge graph embeddings by providing a similarity flooding perspective, proving that existing models seek a fixpoint of pairwise similarities and proposing two effective methods that demonstrate improved performance on benchmark datasets.
Joint representation learning over multi-sourced knowledge graphs (KGs) yields transferable and expressive embeddings that improve downstream tasks. Entity alignment (EA) is a critical step in this process. Despite recent considerable research progress in embedding-based EA, how it works remains to be explored. In this paper, we provide a similarity flooding perspective to explain existing translation-based and aggregation-based EA models. We prove that the embedding learning process of these models actually seeks a fixpoint of pairwise similarities between entities. We also provide experimental evidence to support our theoretical analysis. We propose two simple but effective methods inspired by the fixpoint computation in similarity flooding, and demonstrate their effectiveness on benchmark datasets. Our work bridges the gap between recent embedding-based models and the conventional similarity flooding algorithm. It would improve our understanding of and increase our faith in embedding-based EA.