Scalable Zero-shot Entity Linking with Dense Entity Retrieval

TL;DR

This paper presents a scalable, zero-shot entity linking method using dense retrieval with BERT, achieving state-of-the-art accuracy and speed, and demonstrating effective knowledge transfer between models.

Contribution

Introduces a simple two-stage zero-shot entity linking approach combining dense retrieval and re-ranking, with significant accuracy and speed improvements.

Findings

State-of-the-art zero-shot accuracy on benchmarks

Fast retrieval with 5.9 million candidates in 2 milliseconds

Knowledge distillation enhances bi-encoder performance

Abstract

This paper introduces a conceptually simple, scalable, and highly effective BERT-based entity linking model, along with an extensive evaluation of its accuracy-speed trade-off. We present a two-stage zero-shot linking algorithm, where each entity is defined only by a short textual description. The first stage does retrieval in a dense space defined by a bi-encoder that independently embeds the mention context and the entity descriptions. Each candidate is then re-ranked with a cross-encoder, that concatenates the mention and entity text. Experiments demonstrate that this approach is state of the art on recent zero-shot benchmarks (6 point absolute gains) and also on more established non-zero-shot evaluations (e.g. TACKBP-2010), despite its relative simplicity (e.g. no explicit entity embeddings or manually engineered mention tables). We also show that bi-encoder linking is very fast with nearest neighbour search (e.g. linking with 5.9 million candidates in 2 milliseconds), and that much of the accuracy gain from the more expensive cross-encoder can be transferred to the bi-encoder via knowledge distillation. Our code and models are available at https://github.com/facebookresearch/BLINK.