BriefGPT.xyz
Dec, 2016
通过联合知识增强训练将词语和意义嵌入在一起
Embedding Words and Senses Together via Joint Knowledge-Enhanced Training
HTML
PDF
Massimiliano Mancini, Jose Camacho-Collados, Ignacio Iacobacci, Roberto Navigli
TL;DR
提出一种新模型共同学习单词和语义嵌入，以解决现有方法无法区分单词不同含义的问题，利用大规模预料库和语义网络优化嵌入空间，并在各种任务中评估了该方法的优点。
Abstract
word embeddings
based on neural networks are widely used in
natural language processing
. However, despite their success in capturing
semantic inf
→