EQUATE -EQUATE is to gender bias as Grammarly is to grammar

future implementation

Inspiration

When I first got into natural language processing, I came across some interesting paper titles such as:

"Man is to Computer Programmer as Woman is to Homemaker?"

"King – Man + Woman = Queen: The Marvelous Mathematics of Computational Linguistics"

I learned that the natural language processing behind ChatGPT could be applied to gender bias. Today, social media is filled with a lot of unfriendly content, but if we could integrate this technology into the input methods we use in our daily lives, it could help create a more harmonious community.

reference

Tolga Bolukbasi, Kai-Wei Chang, James Zou, Venkatesh Saligrama, and Adam Kalai. 2016. Man is to computer programmer as woman is to homemaker? debiasing word embeddings. In Proceedings of the 30th International Conference on Neural Information Processing Systems (NIPS'16). Curran Associates Inc., Red Hook, NY, USA, 4356–4364.

https://www.technologyreview.com/2015/09/17/166211/king-man-woman-queen-the-marvelous-mathematics-of-computational-linguistics/

Zhang, G., Li, Y., Wu, Y., Zhang, L., Lin, C., Geng, J., Wang, S., & Fu, J. (2023). CORGI-PM: A Chinese Corpus For Gender Bias Probing and Mitigation. ArXiv, abs/2301.00395.

What it does

EQUATE is to gender bias as Grammarly is to grammar.

Language is a tool of reason and a reflection of the world, and it is bound to be constantly destroyed, evolved, and interacted with the public according to the emergence of new things and phenomena or social changes. Gender neutrality does not mean the neutralization of gender, but the elimination of discrimination from gender distinction and the realization of equality. This is why the linguistic expression is called an inclusive language that encompasses LGBTQ, which is not defined by the male/female dichotomy.I hope that people can transform their language habit from gender-discriminative to gender-equal as the our application let people know that their text have gender-biased contents.

How we built it

We use Language Model ‘api provided by Open AI. Choose Python Django to create the project. Also use PyTorch.

Challenges we ran into

Time limited. So now there is no implementation to integrate Equate into the mobile keyboard input to let users use it everywhere and anytime when they communicate with others.

Accomplishments that we're proud of

None of the three of us share a first language, so we are proud of our ability to communicate regardless of our backgrounds. Also, we are proud to come from different backgrounds, which we believe enhanced our understanding of equity within communication. We are happy with the way that our ideation process went, with lots of intense brainstorming and visual aid. Overall, we are happy with what we have created as we are hopeful that this kind of technology can create a more equitable future.

What we learned

We learned a lot about our own differences in communication styles, and how gender equity is far more applicable to other media besides AI writing. Additionally, we learned about each other and about how we each have almost completely different understandings of the topic we chose. We learned a lot about gender equity in general and are coming out of this feeling a lot more knowledgeable and prepared to talk about gender equity.