Filters








893 Hits in 9.1 sec

The Woman Worked as a Babysitter: On Biases in Language Generation [article]

Emily Sheng, Kai-Wei Chang, Premkumar Natarajan, Nanyun Peng
2019 arXiv   pre-print
In this work, we introduce the notion of the regard towards a demographic, use the varying levels of regard towards different demographics as a defining metric for bias in NLG, and analyze the extent to  ...  We present a systematic study of biases in natural language generation (NLG) by analyzing text generated from prompts that contain mentions of different demographic groups.  ...  Acknowledgments This work was supported by the DARPA UGB program under ISI prime contract HR0011-18-9-0019.  ... 
arXiv:1909.01326v2 fatcat:m6lxcgx3araf5bk36t6wnu34f4

The Woman Worked as a Babysitter: On Biases in Language Generation

Emily Sheng, Kai-Wei Chang, Premkumar Natarajan, Nanyun Peng
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
In this work, we introduce the notion of the regard towards a demographic, use the varying levels of regard towards different demographics as a defining metric for bias in NLG, and analyze the extent to  ...  We present a systematic study of biases in natural language generation (NLG) by analyzing text generated from prompts that contain mentions of different demographic groups.  ...  Acknowledgments This work was supported by the DARPA UGB program under ISI prime contract HR0011-18-9-0019.  ... 
doi:10.18653/v1/d19-1339 dblp:conf/emnlp/ShengCNP19 fatcat:s2fejscoavgnfebk5gz57ykwxe

UnQovering Stereotyping Biases via Underspecified Questions [article]

Tao Li, Tushar Khot, Daniel Khashabi, Ashish Sabharwal, Vivek Srikumar
2020 arXiv   pre-print
We present UNQOVER, a general framework to probe and quantify biases through underspecified questions.  ...  on bias varies strongly with the dataset and the model size.  ...  Acknowledgements We thank Noah Smith, Suresh Venkatasubramanian and Maarten Sap for their valuable insights and suggestions, and also the reviewers and the ethics committee of EMNLP for constructive comments  ... 
arXiv:2010.02428v3 fatcat:s3nao6wikbbvfchrpuumwvnqci

Bias Out-of-the-Box: An Empirical Analysis of Intersectional Occupational Biases in Popular Generative Language Models [article]

Hannah Kirk, Yennie Jun, Haider Iqbal, Elias Benussi, Filippo Volpin, Frederic A. Dreyer, Aleksandar Shtedritski, Yuki M. Asano
2021 arXiv   pre-print
We focus on generative language models as they are well-suited for extracting biases inherited from training data.  ...  Specifically, we conduct an in-depth analysis of GPT-2, which is the most downloaded text generation model on HuggingFace, with over half a million downloads per month.  ...  Acknowledgements We thank the four anonymous reviewers whose suggestions helped improve and clarify this article. We also thank R. Maria del Rio-Chanona and Gesa Biermann for their useful comments.  ... 
arXiv:2102.04130v3 fatcat:ytv3674owzftfl4f7abk27xan4

What do Bias Measures Measure? [article]

Sunipa Dev, Emily Sheng, Jieyu Zhao, Jiao Sun, Yu Hou, Mattie Sanseverino, Jiin Kim, Nanyun Peng, Kai-Wei Chang
2021 arXiv   pre-print
To address this gap, this work presents a comprehensive survey of existing bias measures in NLP as a function of the associated NLP tasks, metrics, datasets, and social biases and corresponding harms.  ...  Natural Language Processing (NLP) models propagate social biases about protected attributes such as gender, race, and nationality.  ...  The woman worked as a babysitter: On biases in language generation.  ... 
arXiv:2108.03362v1 fatcat:migucaqmgne3hiyirl755o4dpu

StylePTB: A Compositional Benchmark for Fine-grained Controllable Text Style Transfer [article]

Yiwei Lyu, Paul Pu Liang, Hai Pham, Eduard Hovy, Barnabás Póczos, Ruslan Salakhutdinov, Louis-Philippe Morency
2021 arXiv   pre-print
Many of the existing style transfer benchmarks primarily focus on individual high-level semantic changes (e.g. positive to negative), which enable controllability at a high level but do not offer fine-grained  ...  As a result, StylePTB brings novel challenges that we hope will encourage future research in controllable text style transfer, compositional models, and learning disentangled representations.  ...  HP and BP are supported by the DARPA D3M Program and The Boeing Company. RS was supported in part by NSF IIS1763562 and ONR Grant N000141812861.  ... 
arXiv:2104.05196v1 fatcat:2wiad5uainbblbyzhb4wz5rhrq

RedditBias: A Real-World Resource for Bias Evaluation and Debiasing of Conversational Language Models [article]

Soumya Barikeri, Anne Lauscher, Ivan Vulić, Goran Glavaš
2021 arXiv   pre-print
Recent work has predominantly focused on measuring and mitigating bias in pretrained language models.  ...  resources, and completely ignores the impact that debiasing methods may have on the final performance in dialog tasks, e.g., conversational response generation.  ...  The woman worked as a babysitter: On biases in language generation.  ... 
arXiv:2106.03521v1 fatcat:s43dffsi25goppqnb742qxcf7e

Affectations, Conformation and Defiance in Americanah by Chimamanda Ngozi Adichie

Abdullah Nejat TÖNGÜR
2021 Dil ve Edebiyat Araştırmaları/Journal of Language and Literature Studies  
Chimamanda Ngozi Adichie's third novel Americanah portrays a young Nigerian woman, Ifemelu, who goes to the USA for university education, stays on in the country for 13 years and returns to Nigeria despite  ...  With alternate perspectives of Ifemelu, the protagonist, and her love, Obinze, who lives in the UK as an illegal immigrant for three years, Adichie explores the issue of being a foreigner in another country  ...  Aunty Uju's life as a mistress of a Nigerian general, The General, her subservient position in her relationship with him and her financial dependency on the man despite her profession as a medical doctor  ... 
doi:10.30767/diledeara.899574 fatcat:nmbqcnwdczh4xat4nf3eb7mitq

Reclaiming the Subject: Italian Women Self-Defined

Chris Ruggiero
1986 Explorations in ethnic studies : the journal of the National Association of Interdisciplinary Ethnic Studies  
and competent in her role as a woman  ...  becomes the woman as wage laborer, housekeeper, babysitter, tutor, laundress, gardener, chauffeur and cook.  ... 
doi:10.1525/ees.1986.9.1.27 fatcat:bjwgzy3gurexjenidibfgbdzrq

Diversity and power in the English hotel conversations textbook

Ali Roohani, Anahita Sheikhipour, Maryam Farhang
2020 International Journal of Research Studies in Education  
The study, thus, investigated the visual representation of gender in power positions and examined the frequency of the ethnic groups in power positions in English Hotel Conversations, a textbook on hotel  ...  Only a small number of pictures represented the people from various ethnic groups.  ...  In this research, five strategies were introduced to tackle with cultural biases and stereotypic attitudes in English as a second language (ESL) materials.  ... 
doi:10.5861/ijrse.2020.5048 fatcat:pp5333kfabhzblihwoken6g77a

Mother Like Mothers and Work Like Fathers: U.S. Heterosexual College Students' Assumptions About Who Should Meet Childcare and Housework Demands

Annie McConnon, Allegra J. Midgette, Clare Conry-Murray
2021 Sex Roles  
attitudes than in the past.  ...  Results revealed that most participants wanted mothers to have the choice to work but considered childcare a limiting problem that (primarily) mothers should solve.  ...  The writing of this manuscript was supported in part by a postdoctoral fellowship provided by the National Institute of Child Health and Human Development (T32-HD007376) through the Frank Porter Graham  ... 
doi:10.1007/s11199-021-01252-3 pmid:34725534 pmcid:PMC8551347 fatcat:mk4xrzyn25bjvjgw6r2vyd6rba

"You are grounded!": Latent Name Artifacts in Pre-trained Language Models [article]

Vered Shwartz, Rachel Rudinger, Oyvind Tafjord
2020 arXiv   pre-print
Pre-trained language models (LMs) may perpetuate biases originating in their training corpus to downstream models.  ...  As a silver lining, our experiments suggest that additional pre-training on different corpora may mitigate this bias.  ...  In AAAI. Emily Sheng, Kai-Wei Chang, Premkumar Natarajan, and Nanyun Peng. 2019. The woman worked as a babysitter: On biases in language generation.  ... 
arXiv:2004.03012v2 fatcat:fw4gsqbffvbbxenarrdy6xtkyy

Gendered Word (or World): Sexism in Philippine Preschool English Language Textbooks

Veronico Nogales Tarrayo
2014 i-manager's Journal on English Language Teaching  
For instance, in the area of appearance, there are approved patterns for grooming, hairstyles, and clothing for the two sexes.  ...  Clothes that symbolize abilities to do things are regarded as appropriate for males and those that symbolize dependency and inability to walk long distances because  ...  Here are some examples: • Generic masculine pronoun (Every student has to submit his project); • Word connotations (call boy, call actors before they go on the stage versus call girl, a prostitute; woman  ... 
doi:10.26634/jelt.4.2.2795 fatcat:ewy6ex4ynvhatnpiybbopgp2eq

BBQ: A Hand-Built Bias Benchmark for Question Answering [article]

Alicia Parrish, Angelica Chen, Nikita Nangia, Vishakh Padmakumar, Jason Phang, Jana Thompson, Phu Mon Htut, Samuel R. Bowman
2022 arXiv   pre-print
It is well documented that NLP models learn social biases, but little work has been done on how these biases manifest in model outputs for applied tasks like question answering (QA).  ...  We find that models often rely on stereotypes when the context is under-informative, meaning the model's outputs consistently reproduce harmful biases in this setting.  ...  This material is based upon work supported by the National Science Foundation under Grant Nos. 1922658 and 2046556.  ... 
arXiv:2110.08193v2 fatcat:glxryixd7fecfhflsxp356ey2y

Grammatical Gender Trouble and Hungarian Gender[lessness]. Part I: Comparative Linguistic Gender

Louise O. Vasvári
2011 Hungarian Cultural Studies  
, with only pronominal gender (sometimes misnamed 'natural gender'), and languages such as Hungarian and other Finno-Ugric languages, as well as many other languages in the world, such as Turkish and Chinese  ...  While in a narrow linguistic sense linguistic gender can be said to be afunctional, this does not take into account the ideological ramifications in gendered languages of the "leakage" between gender and  ...  Or as she put it: "Gender is the enforcement of sex in language, working in the same way as the declaration of sex in civil status."  ... 
doi:10.5195/ahea.2011.40 fatcat:jpdyyseqcfg3bc5zkyjabliglm
« Previous Showing results 1 — 15 out of 893 results