BriefGPT.xyz
Apr, 2019
使用多视图变分自编码器结合情感词典
Combining Sentiment Lexica with a Multi-View Variational Autoencoder
HTML
PDF
Alexander Hoyle, Lawrence Wolf-Sonkin, Hanna Wallach, Ryan Cotterell, Isabelle Augenstein
TL;DR
介绍了一种生成情感词汇表的模型——SentiVAE，它能够将不同比例尺下的标签统一到一个通用的潜在表征上，并将其应用于情感文本分类任务中并取得了成功。
Abstract
When assigning
quantitative labels
to a dataset, different methodologies may rely on different scales. In particular, when assigning polarities to words in a
sentiment lexicon
, annotators may use binary, categori
→