Complex and Holographic Embeddings of Knowledge Graphs: A Comparison

Abstract

Embeddings of knowledge graphs have received significant attention due to their excellent performance for tasks like link prediction and entity resolution. In this short paper, we are providing a comparison of two state-of-the-art knowledge graph embeddings for which their equivalence has recently been established, i.e., ComplEx and HolE [Nickel, Rosasco, and Poggio, 2016; Trouillon et al., 2016; Hayashi and Shimbo, 2017]. First, we briefly review both models and discuss how their scoring functions are equivalent. We then analyze the discrepancy of results reported in the original articles, and show experimentally that they are likely due to the use of different loss functions. In further experiments, we evaluate the ability of both models to embed symmetric and antisymmetric patterns. Finally, we discuss advantages and disadvantages of both models and under which conditions one would be preferable to the other.

References in corpus (4)

• Complex Embeddings for Simple Link Prediction
• Deep Complex Networks
• Knowledge Graph Completion via Complex Tensor Factorization
• On the Equivalence of Holographic and Complex Embeddings for Link Prediction

Cited by in corpus (10)

• Quaternion Knowledge Graph Embeddings
• Deep Complex Networks
• Knowledge Graph Completion via Complex Tensor Factorization
• Fast Linear Model for Knowledge Graph Embeddings
• A Complex KBQA System using Multiple Reasoning Paths
• Knowledge Graph Embedding with Multiple Relation Projections
• WikiDataSets: Standardized sub-graphs from Wikidata
• Out-of-Vocabulary Entities in Link Prediction
• A Relational Tucker Decomposition for Multi-Relational Link Prediction
• Knowledge Sheaves: A Sheaf-Theoretic Framework for Knowledge Graph Embedding