Summary of Gender Bias Mitigation For Bangla Classification Tasks, by Sajib Kumar Saha Joy et al.
Gender Bias Mitigation for Bangla Classification Tasks
by Sajib Kumar Saha Joy, Arman Hassan Mahy, Meherin Sultana, Azizah Mamun Abha, MD Piyal Ahmmed, Yue Dong, G M Shahariar
First submitted to arxiv on: 16 Nov 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The study investigates gender bias in Bangla pretrained language models, a largely underexplored area in low-resource languages. Researchers applied gender-name swapping techniques to existing datasets to create four manually annotated datasets for sentiment analysis, toxicity detection, hate speech detection, and sarcasm detection. The goal was to detect and mitigate gender bias using task-specific pretrained models. A joint loss optimization technique was proposed to reduce gender bias across tasks while maintaining competitive accuracy compared to other baseline approaches. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The study looks at how language models can be biased towards one gender over another in a particular language, Bangla. To see if this is true, they took existing datasets and changed the names and words related to gender to make them more neutral. This created four new datasets that could be used to test and fix this bias. They came up with a way to make the models less biased while still being good at understanding what’s in texts. |
Keywords
* Artificial intelligence * Optimization
