Unsupervised Distillation of Syntactic Information from Contextualized Word Representations

by   Shauli Ravfogel, et al.

Contextualized word representations, such as ELMo and BERT, were shown to perform well on various semantic and syntactic tasks. In this work, we tackle the task of unsupervised disentanglement between semantics and structure in neural language representations: we aim to learn a transformation of the contextualized vectors, that discards the lexical semantics, but keeps the structural information. To this end, we automatically generate groups of sentences which are structurally similar but semantically different, and use metric-learning approach to learn a transformation that emphasizes the structural component that is encoded in the vectors. We demonstrate that our transformation clusters vectors in space by structural properties, rather than by lexical semantics. Finally, we demonstrate the utility of our distilled representations by showing that they outperform the original contextualized representations in a few-shot parsing setting.



There are no comments yet.


page 1

page 2

page 3

page 4


Frame-Based Continuous Lexical Semantics through Exponential Family Tensor Factorization and Semantic Proto-Roles

We study how different frame annotations complement one another when lea...

Parsing using a grammar of word association vectors

This paper was was first drafted in 2001 as a formalization of the syste...

Analysing Lexical Semantic Change with Contextualised Word Representations

This paper presents the first unsupervised approach to lexical semantic ...

How Language-Neutral is Multilingual BERT?

Multilingual BERT (mBERT) provides sentence representations for 104 lang...

Syn-QG: Syntactic and Shallow Semantic Rules for Question Generation

Question Generation (QG) is fundamentally a simple syntactic transformat...

Personalized word representations Carrying Personalized Semantics Learned from Social Network Posts

Distributed word representations have been shown to be very useful in va...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.