ChatDEI

Our machine learning model show a very high accuracy

Inspiration

We are a group of high school students from Toronto, Canada who are passionate about Diversity, Equity, and Inclusion (DEI). As racial minorities ourselves, we have witnessed and experienced the impact of harmful stereotypes on marginalized groups, both online and in real life. To break down the barrier among different groups, we must tackle the challenge of recognizing and correcting assumptions based on race, gender, age, education, income, mental health, and language.

What it does

ChatDEI is more than just a tool; it is a catalyst for change. Enter any text, and our Machine Learning model will detect the presence of offensive language and stereotypes with high accuracy. Then, ChatDEI provides suggestions for correction, prompting the users to reflect on their communication choices in the meantime.

How we built it

This technology is built using HTML, CSS, Python (Google Colab, Flask), and most importantly, the convolutional neural network (CNN).

Challenges we ran into

1. How to build a website using HTML and CSS Our team learned how to build a multi-page, responsive, and aesthetically-pleasing website in three days.

2 .Connecting the back-end with the front-end We tried numerous methods to connect the back with the front, eventually settling down on Google Colab.

3. Project organization Originally, we planned to work on GitHub codespaces. However, this site would not allow the three of us to collaborate in the same space. Therefore, the front-end developers switched to Replit, and the back-end programmers worked with Flask and Google Collab.