Evaluating Design Decisions for Dual Encoder-based Entity Disambiguation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Rücker, Susanna, Akbik, Alan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909614347386880
author Rücker, Susanna
Akbik, Alan
author_facet Rücker, Susanna
Akbik, Alan
contents Entity disambiguation (ED) is the task of linking mentions in text to corresponding entries in a knowledge base. Dual Encoders address this by embedding mentions and label candidates in a shared embedding space and applying a similarity metric to predict the correct label. In this work, we focus on evaluating key design decisions for Dual Encoder-based ED, such as its loss function, similarity metric, label verbalization format, and negative sampling strategy. We present the resulting model VerbalizED, a document-level Dual Encoder model that includes contextual label verbalizations and efficient hard negative sampling. Additionally, we explore an iterative prediction variant that aims to improve the disambiguation of challenging data points. Comprehensive experiments on AIDA-Yago validate the effectiveness of our approach, offering insights into impactful design choices that result in a new State-of-the-Art system on the ZELDA benchmark.
format Preprint
id arxiv_https___arxiv_org_abs_2505_11683
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Evaluating Design Decisions for Dual Encoder-based Entity Disambiguation
Rücker, Susanna
Akbik, Alan
Computation and Language
Entity disambiguation (ED) is the task of linking mentions in text to corresponding entries in a knowledge base. Dual Encoders address this by embedding mentions and label candidates in a shared embedding space and applying a similarity metric to predict the correct label. In this work, we focus on evaluating key design decisions for Dual Encoder-based ED, such as its loss function, similarity metric, label verbalization format, and negative sampling strategy. We present the resulting model VerbalizED, a document-level Dual Encoder model that includes contextual label verbalizations and efficient hard negative sampling. Additionally, we explore an iterative prediction variant that aims to improve the disambiguation of challenging data points. Comprehensive experiments on AIDA-Yago validate the effectiveness of our approach, offering insights into impactful design choices that result in a new State-of-the-Art system on the ZELDA benchmark.
title Evaluating Design Decisions for Dual Encoder-based Entity Disambiguation
topic Computation and Language
url https://arxiv.org/abs/2505.11683