Hyperbolic Entailment Cones for Learning Hierarchical Embeddings

Abstract

Learning graph representations via low-dimensional embeddings that preserve relevant network properties is an important class of problems in machine learning. We here present a novel method to embed directed acyclic graphs. Following prior work, we first advocate for using hyperbolic spaces which provably model tree-like structures better than Euclidean geometry. Second, we view hierarchical relations as partial orders defined using a family of nested geodesically convex cones. We prove that these entailment cones admit an optimal shape with a closed form expression both in the Euclidean and hyperbolic spaces, and they canonically define the embedding learning process. Experiments show significant improvements of our method over strong recent baselines both in terms of representational capacity and generalization.

arXiv commentInternational Conference on Machine Learning (ICML) 2018

References in corpus (6)

• Distributed Representations of Words and Phrases and their Compositionality
• Geometric deep learning: going beyond Euclidean data
• Graph Embedding Techniques, Applications, and Performance: A Survey
• Hyperbolic Geometry of Complex Networks
• Reasoning about Entailment with Neural Attention
• Order-Embeddings of Images and Language

Cited by in corpus (14)

• Discriminative Topic Mining via Category-Name Guided Text Embedding
• Hyperbolic Neural Networks
• Riemannian adaptive stochastic gradient algorithms on matrix manifolds
• Scalable Hyperbolic Recommender Systems
• Hyperbolic Graph Attention Network
• Hyperbolic Interaction Model For Hierarchical Multi-Label Classification
• Symmetric Spaces for Graph Embeddings: A Finsler-Riemannian Approach
• Hyperbolic Deep Learning for Foundation Models: A Survey
• Probabilistic Box Embeddings for Uncertain Knowledge Graph Reasoning
• Embedding Heterogeneous Networks into Hyperbolic Space Without Meta-path
• Multi-modal Entity Alignment in Hyperbolic Space
• Exploring Semantic Capacity of Terms
• Hyperbolic Embeddings for Learning Options in Hierarchical Reinforcement Learning
• Understanding in Artificial Intelligence