A Causal Inference Method for Reducing Gender Bias in Word Embedding Relations

被引:0
作者
Yang, Zekun [1 ]
Feng, Juan [1 ]
机构
[1] City Univ Hong Kong, Dept Informat Syst, Coll Business, Hong Kong, Peoples R China
来源
THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE | 2020年 / 34卷
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Word embedding has become essential for natural language processing as it boosts empirical performances of various tasks. However, recent research discovers that gender bias is incorporated in neural word embeddings, and downstream tasks that rely on these biased word vectors also produce gender-biased results. While some word-embedding gender-debiasing methods have been developed, these methods mainly focus on reducing gender bias associated with gender direction and fail to reduce the gender bias presented in word embedding relations. In this paper, we design a causal and simple approach for mitigating gender bias in word vector relation by utilizing the statistical dependency between gender-definition word embeddings and gender-biased word embeddings. Our method attains state-of-the-art results on gender-debiasing tasks, lexical- and sentence-level evaluation tasks, and downstream coreference resolution tasks.
引用
收藏
页码:9434 / 9441
页数:8
相关论文
共 34 条
[1]  
Agirre E., 2015, P 9 INT WORKSH SEM E, P252, DOI [10.18653/v1/S15-2045, DOI 10.18653/V1/S15-2045]
[2]  
Agirre E., 2014, P 8 INT WORKSH SEM E, P81, DOI [10.3115/v1/S14-2010, DOI 10.3115/V1/S14-2010]
[3]  
Agirre Eneko, 2013, P 2 JOINT C LEX COMP, P32
[4]  
Agirre Eneko, 2012, SEM 2012 1 JOINT C L, P385
[5]  
[Anonymous], 2011, P 20 INT C WORLD WID
[6]  
[Anonymous], 2018, P 2018 C EMP METH NA
[7]  
[Anonymous], 2008, Introduction to information retrieval
[8]  
[Anonymous], 1995, RES DESIGN STAT ANAL
[9]  
Bolukbasi T, 2016, ADV NEUR IN, V29
[10]   Multimodal Distributional Semantics [J].
Bruni, Elia ;
Nam Khanh Tran ;
Baroni, Marco .
JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH, 2014, 49 :1-47