Causally Denoise Word Embeddings Using Half-Sibling Regression

Research output: Chapters, Conference Papers, Creative and Literary Works (RGC: 12, 32, 41, 45)32_Refereed conference paper (with ISBN/ISSN)peer-review

1 Scopus Citations
View graph of relations

Author(s)

Detail(s)

Original languageEnglish
Title of host publicationThe Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI-20)
Place of PublicationCalifornia
PublisherAAAI Press
Pages9426-9433
ISBN (Print)9781577358350 (set)
Publication statusPublished - Feb 2020

Publication series

NameProceedings of the AAAI Conference on Artificial Intelligence
PublisherAAAI Press
Number5
Volume34
ISSN (Print)2159-5399
ISSN (Electronic)2374-3468

Conference

Title34th AAAI Conference on Artificial Intelligence (AAAI-20)
PlaceUnited States
CityNew York
Period7 - 12 February 2020

Abstract

Distributional representations of words, also known as word vectors, have become crucial for modern natural language processing tasks due to their wide applications. Recently, a growing body of word vector postprocessing algorithm has emerged, aiming to render off-the-shelf word vectors even stronger. In line with these investigations, we introduce a novel word vector postprocessing scheme under a causal inference framework. Concretely, the postprocessing pipeline is realized by Half-Sibling Regression (HSR), which allows us to identify and remove confounding noise contained in word vectors. Compared to previous work, our proposed method has the advantages of interpretability and transparency due to its causal inference grounding. Evaluated on a battery of standard lexical-level evaluation tasks and downstream sentiment analysis tasks, our method reaches state-of-the-art performance.

Citation Format(s)

Causally Denoise Word Embeddings Using Half-Sibling Regression. / Yang, Zekun; Liu, Tianlin.

The Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI-20). California : AAAI Press, 2020. p. 9426-9433 (Proceedings of the AAAI Conference on Artificial Intelligence; Vol. 34, No. 5).

Research output: Chapters, Conference Papers, Creative and Literary Works (RGC: 12, 32, 41, 45)32_Refereed conference paper (with ISBN/ISSN)peer-review