BriefGPT.xyz
Aug, 2016
用于概率神经词嵌入的形态学先验
Morphological Priors for Probabilistic Neural Word Embeddings
HTML
PDF
Parminder Bhatia, Robert Guthrie, Jacob Eisenstein
TL;DR
通过将形态学信息融合到词向量中，构建了一个统一的概率框架，其中词嵌入是一个潜变量，并以形态学信息提供先验分布。此方法改进了内在词相似性评估，也在词性标注下游任务中提高了性能。
Abstract
word embeddings
allow natural language processing systems to share statistical information across related words. These embeddings are typically based on
distributional statistics
, making it difficult for them to
→