Dictionary-based Debiasing of Pre-trained Word Embeddings

01/23/2021
by   Masahiro Kaneko, et al.
0

Word embeddings trained on large corpora have shown to encode high levels of unfair discriminatory gender, racial, religious and ethnic biases. In contrast, human-written dictionaries describe the meanings of words in a concise, objective and an unbiased manner. We propose a method for debiasing pre-trained word embeddings using dictionaries, without requiring access to the original training resources or any knowledge regarding the word embedding algorithms used. Unlike prior work, our proposed method does not require the types of biases to be pre-defined in the form of word lists, and learns the constraints that must be satisfied by unbiased word embeddings automatically from dictionary definitions of the words. Specifically, we learn an encoder to generate a debiased version of an input word embedding such that it (a) retains the semantics of the pre-trained word embeddings, (b) agrees with the unbiased definition of the word according to the dictionary, and (c) remains orthogonal to the vector space spanned by any biased basis vectors in the pre-trained word embedding space. Experimental results on standard benchmark datasets show that the proposed method can accurately remove unfair biases encoded in pre-trained word embeddings, while preserving useful semantics.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
06/03/2019

Gender-preserving Debiasing for Pre-trained Word Embeddings

Word embeddings learnt from massive text collections have demonstrated s...
research
10/25/2020

Autoencoding Improves Pre-trained Word Embeddings

Prior work investigating the geometry of pre-trained word embeddings hav...
research
01/23/2021

Debiasing Pre-trained Contextualised Embeddings

In comparison to the numerous debiasing methods proposed for the static ...
research
03/14/2020

Word Sense Disambiguation for 158 Languages using Word Embeddings Only

Disambiguation of word senses in context is easy for humans, but is a ma...
research
07/21/2017

An Error-Oriented Approach to Word Embedding Pre-Training

We propose a novel word embedding pre-training approach that exploits wr...
research
10/18/2022

On the Information Content of Predictions in Word Analogy Tests

An approach is proposed to quantify, in bits of information, the actual ...
research
05/27/2022

Semeval-2022 Task 1: CODWOE – Comparing Dictionaries and Word Embeddings

Word embeddings have advanced the state of the art in NLP across numerou...

Please sign up or login with your details

Forgot password? Click here to reset