Domain Adaptation Multi-task Deep Neural Network for Mitigating Unintended Bias in Toxic Language Detection

Farshid Faal, Farshid Faal, Jia Yu, Jia Yu, Ketra Schmitt, Ketra Schmitt

2021

Abstract

As online communities have grown, so has the ability to exchange ideas, which includes an increase in the spread of toxic language, including racism, sexual harassment, and other negative behaviors that are not tolerated in polite society. Hence, toxic language detection within online conversations has become an essential application of natural language processing. In recent years, machine learning approaches for toxic language detection have primarily focused on many researchers in academics and industries. However, in many of these machine learning models, non-toxic comments containing specific identity terms, such as gay, Black, Muslim, and Jewish, were given unreasonably high toxicity scores. In this research, we propose a new approach based on the domain adaptation language model and multi-task deep neural network to identify and mitigate this form of unintended model bias in online conversations. We use six toxic language detection and identification tasks to train the model to detect toxic contents and mitigate unintended bias in model prediction. We evaluate our model and compare it with other state-of-the-art deep learning models using specific performance metrics to measure the model bias. In detailed experiments, we show our approach can identify the toxic language in conversations with considerably more robustness to model bias towards commonly-attacked identity groups presented in online conversations in social media.

Paper Citation

in Harvard Style

Faal F., Yu J. and Schmitt K. (2021). Domain Adaptation Multi-task Deep Neural Network for Mitigating Unintended Bias in Toxic Language Detection.In Proceedings of the 13th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART, ISBN 978-989-758-484-8, pages 932-940. DOI: 10.5220/0010266109320940

in Bibtex Style

@conference{icaart21,
author={Farshid Faal and Jia Yu and Ketra Schmitt},
title={Domain Adaptation Multi-task Deep Neural Network for Mitigating Unintended Bias in Toxic Language Detection},
booktitle={Proceedings of the 13th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,},
year={2021},
pages={932-940},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010266109320940},
isbn={978-989-758-484-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 13th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART,
TI - Domain Adaptation Multi-task Deep Neural Network for Mitigating Unintended Bias in Toxic Language Detection
SN - 978-989-758-484-8
AU - Faal F.
AU - Yu J.
AU - Schmitt K.
PY - 2021
SP - 932
EP - 940
DO - 10.5220/0010266109320940