# Fine-Grained Entity Typing in Hyperbolic Space

## TL;DR

This paper explores the use of hyperbolic embeddings to better capture hierarchical relations in large entity type inventories for fine-grained entity typing, comparing their effectiveness to Euclidean models across different datasets.

## Contribution

It demonstrates the potential of hyperbolic space for modeling hierarchical entity types and analyzes factors affecting its performance in entity typing tasks.

## Key findings

- Hyperbolic embeddings outperform Euclidean in some cases.
- Effectiveness depends on type inventory granularity.
- Hierarchical relation inference impacts model performance.

## Abstract

How can we represent hierarchical information present in large type inventories for entity typing? We study the ability of hyperbolic embeddings to capture hierarchical relations between mentions in context and their target types in a shared vector space. We evaluate on two datasets and investigate two different techniques for creating a large hierarchical entity type inventory: from an expert-generated ontology and by automatically mining type co-occurrences. We find that the hyperbolic model yields improvements over its Euclidean counterpart in some, but not all cases. Our analysis suggests that the adequacy of this geometry depends on the granularity of the type inventory and the way hierarchical relations are inferred.