BiasDetective

BiasDetective Main Page

Inspiration

My mom has dealt a lot with sexism in her search for jobs and after accepting jobs as well. Based on the prompts given and my moms experience I wanted to create an app that can detect the possibility of bias, so that the user can determine whether or not the employer is biased.

What it does

The user enters a piece of text and it is then evaluated for bias and neutrality. It shows a bias percentage and Neutrality percentage. So the bias percentage determines the likelihood the text is biased and the neutrality percentage determines the likelihood the text is gender-neutral.

How we built it

I used expo and react native, and I used a Hugging Face model bert-base-uncased-md_gender_bias-trained by JakobKaiser. I used a python programme to determine the scores. I then used ngrok for the domain to be public and then pushed the code to Git Hub.

Challenges we ran into

I wanted to write a programme to highlight sentences with gender bias but with my somewhat limited coding knowledge, I couldn't quite crack it without running into some error.

Accomplishments that we're proud of

This is the first time I've integrated a model into an app or programme so I'm happy its something I was able to learn and I plan to use it in the future.