What relations are reliably embeddable in Euclidean space?

2019/03/13 by Bhattacharjee, Robi, Dasgupta, Sanjoy · 2 citations
#FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML)

paper · doi:10.48550/arxiv.1903.05347

Abstract

We consider the problem of embedding a relation, represented as a directed graph, into Euclidean space. For three types of embeddings motivated by the recent literature on knowledge graphs, we obtain characterizations of which relations they are able to capture, as well as bounds on the minimal dimensionality and precision needed.

Cited by

• Relation Geometry in Semantic Space of Language Models
• Compressibility Barriers to Neighborhood-Preserving Data Visualizations

Related

• Geometric Relational Embeddings: A Survey
• SpaceE: Knowledge Graph Embedding by Relational Linear Transformation in the Entity Space
• Embedding products of graphs into Euclidean spaces
• Interpreting Knowledge Graph Relation Representation from Word Embeddings
• From Knowledge Graph Embedding to Ontology Embedding? An Analysis of the Compatibility between Vector Space Representations and Rules