BriefGPT.xyz
Jan, 2020
命名实体识别中的上下文嵌入：一项关于泛化的实证研究
Contextualized Embeddings in Named-Entity Recognition: An Empirical Study on Generalization
HTML
PDF
Bruno Taillé, Vincent Guigue, Patrick Gallinari
TL;DR
本文通过经验分析，将提到的未知名称根据新奇性和域外评估进行分离，并证明了最新情境嵌入模型在发现前所未见的名称方面特别有益，尤其是在域外评估中。
Abstract
contextualized embeddings
use unsupervised language model pretraining to compute word representations depending on their context. This is intuitively useful for
generalization
, especially in
→