Representing Hierarchical Structure by Using Cone Embedding

TL;DR

This paper introduces cone embedding, a novel method for representing hierarchical structures in graphs within a metric cone, addressing limitations of previous isometric embeddings like Poincaré, and providing intuitive hierarchical indicators.

Contribution

The paper proposes cone embedding in a metric cone, offering a natural hierarchical indicator and enabling extraction of hierarchy from other graph embeddings.

Findings

Provides a geometrically intuitive hierarchy indicator

Enables extraction of hierarchy from existing embeddings

Addresses limitations of isometric mappings in prior methods

Abstract

Graph embedding is becoming an important method with applications in various areas, including social networks and knowledge graph completion. In particular, Poincar\'e embedding has been proposed to capture the hierarchical structure of graphs, and its effectiveness has been reported. However, most of the existing methods have isometric mappings in the embedding space, and the choice of the origin point can be arbitrary. This fact is not desirable when the distance from the origin is used as an indicator of hierarchy, as in the case of Poincar\'e embedding. In this paper, we propose cone embedding, embedding method in a metric cone, which solve these problems, and we gain further benefits: 1) we provide an indicator of hierarchical information that is both geometrically and intuitively natural to interpret, and 2) we can extract the hierarchical structure from a graph embedding output of other methods by learning additional one-dimensional parameters.