Abstract
AI and BERT (Bidirectional Encoder Representations from Transformers) have been increasingly adopted in the human resources (HR) industry for recruitment. The increased efficiency (e.g., fairness) will help remove biases in machine learning, help organizations find a qualified candidate, and remove bias in the labor market. BERT has further improved the performance of language representation models by using an auto-encoding model which incorporates larger bidirectional contexts. However, BERT’s underlying mechanisms that enhance its effectiveness, such as tokenization, masking, and leveraging the attention mechanism to compute vector score, are not well understood.
This research analyzes how BERT’s architecture and its tokenization protocol affect the low number of occurrences of the minority-related data using the cosine similarity of its embeddings. In this project, by using a dataset of racially and gender-associated personal names and analyzing the interactions of transformers, we present the unfair prejudice of BERTs’ pre-trained network and autoencoding model. Furthermore, by analyzing the distance of an initial word’s token and its MASK replacement token using the cosine similarity, we will demonstrate the inherent discriminability during pre-training. Finally, this research will deliver potential solutions to mitigate discrimination and bias in BERT by examining its geometric properties.
Supported by organization x.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bartl, M., Nissim, M., Gatt, A.: Unmasking contextual stereotypes: measuring and mitigating Bert’s gender bias. arXiv preprint arXiv:2010.14534 (2020)
Basta, C., Costa-Jussà, M.R., Casas, N.: Evaluating the underlying gender bias in contextualized word embeddings. arXiv preprint arXiv:1904.08783 (2019)
Bertrand, M., Mullainathan, S.: Replication data for: are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination (2019)
Bhardwaj, R., Majumder, N., Poria, S.: Investigating gender bias in Bert. arXiv preprint arXiv:2009.05021 (2020)
Brunet, M.-E., Alkalay-Houlihan, C., Anderson, A., Zemel, R.: Understanding the origins of bias in word embeddings. In: International Conference on Machine Learning, pp. 803–811. PMLR (2019)
Caliskan, A., Bryson, J.J., Narayanan, A.: Semantics derived automatically from language corpora contain human-like biases. Science 356(6334), 183–186 (2017)
Chaloner, K., Maldonado, A.: Measuring gender bias in word embeddings across domains and discovering new gender bias word categories. In: Proceedings of the 1st Workshop on Gender Bias in Natural Language Processing, pp. 25–32 (2019)
Chang, K.-W., Prabhakaran, V., Ordonez, V.: Bias and fairness in natural language processing. In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP): Tutorial Abstracts (2019)
Church, K.W.: Word2vec. Nat. Lang. Eng. 23(1), 155–162 (2017)
Ontario Human Rights Commission: Racism and racial discrimination: Systemic discrimination (fact sheet). Ontario Human Rights Commission. Accessed 17 July 2018
Devlin, J.: Question: why Bert underperforms issue #80 ukplab/sentence-transformers (November 2018)
Devlin, J., Chang, M.-W., Lee, K., Toutanova, K.: Bert: pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805 (2018)
Hangartner, D., Kopp, D., Siegenthaler, M.: Monitoring hiring discrimination through online recruitment platforms. Nature 589(7843), 572–576 (2021)
Howard, J., Ruder, S.: Universal language model fine-tuning for text classification. arXiv preprint arXiv:1801.06146 (2018)
Kiritchenko, S., Mohammad, S.M.: Examining gender and race bias in two hundred sentiment analysis systems. arXiv preprint arXiv:1805.04508 (2018)
Kurita, K., Vyas, N., Pareek, A., Black, A.W., Tsvetkov, Y.: Measuring bias in contextualized word representations. arXiv preprint arXiv:1906.07337 (2019)
Liddy, E.D.: Natural Language Processing (2001)
Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., Galstyan, A.: A survey on bias and fairness in machine learning. arXiv preprint arXiv:1908.09635 (2019)
Mozafari, M., Farahbakhsh, R., Crespi, N.: Hate speech detection and racial bias mitigation in social media based on Bert model. PLoS ONE 15(8), e0237861 (2020)
Bureau of Labor Statistics: Employed persons by detailed occupation, sex, race, and Hispanic or Latino ethnicity. U.S. Bureau of Labor Statistics (January 2021)
Olteanu, A., Castillo, C., Diaz, F., Kıcıman, E.: Social data: biases, methodological pitfalls, and ethical boundaries. Front. Big Data 2, 13 (2019)
Pennington, J., Socher, R., Manning, C.D.: GloVe: global vectors for word representation. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 1532–1543 (2014)
Peters, M.E., et al.: Deep contextualized word representations. arXiv preprint arXiv:1802.05365 (2018)
Podkorytov, M., Biś, D., Cai, J., Amirizirtol, K., Liu, X.: Effects of architecture and training on embedding geometry and feature discriminability in Bert. In: 2020 International Joint Conference on Neural Networks (IJCNN), pp. 1–8. IEEE (2020)
Harward Project Implicit. https://www.projectimplicit.net
Reimers, N.: Why cosine similarity of BERT, ALBERT, Robert is so big, almost near 1.0? \(\cdot \) issue #2298 \(\cdot \) huggingface/transformers (December 2019)
Sharma, S., Dey, M., Sinha, K.: Evaluating gender bias in natural language inference (2021)
Steed, R., Caliskan, A.: Image representations learned with unsupervised pre-training contain human-like biases. arXiv preprint arXiv:2010.15052 (2020)
Suresh, H., Guttag, J.V.: A framework for understanding unintended consequences of machine learning. arXiv preprint arXiv:1901.10002 (2019)
Wu, Y., et al.: Google’s neural machine translation system: bridging the gap between human and machine translation. arXiv preprint arXiv:1609.08144 (2016)
Zhang, L., Wu, Y., Wu, X.: A causal framework for discovering and removing direct and indirect discrimination. arXiv preprint arXiv:1611.07509 (2016)
Zhao, J., Wang, T., Yatskar, M., Cotterell, R., Ordonez, V., Chang, K.-W.: Gender bias in contextualized word embeddings. In: Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), Minneapolis, Minnesota, June 2019, pp. 629–634. Association for Computational Linguistics (2019)
Zhu, Y., et al.: Aligning books and movies: towards story-like visual explanations by watching movies and reading books. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 19–27 (2015)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Ramezanzadehmoghadam, M., Chi, H., Jones, E.L., Chi, Z. (2021). Inherent Discriminability of BERT Towards Racial Minority Associated Data. In: Gervasi, O., et al. Computational Science and Its Applications – ICCSA 2021. ICCSA 2021. Lecture Notes in Computer Science(), vol 12951. Springer, Cham. https://doi.org/10.1007/978-3-030-86970-0_19
Download citation
DOI: https://doi.org/10.1007/978-3-030-86970-0_19
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-86969-4
Online ISBN: 978-3-030-86970-0
eBook Packages: Computer ScienceComputer Science (R0)