Inspiration
Athletes and celebrities are asked very different comments. Females are asked more about their clothes and fashion choices while men are often asked about their skills. This got us thinking about how there are differences in how women are discriminated against even now. Although it started with visual and audio media, as we dug deeper, we found more the written medium far more interesting.
What it does
You copy the article or excerpt that you want to check for bias and sentimentality, then you can easily observe the different results that come out of the app by scrolling down.
How we built it
Put simply, it is a Web based application built with a JavaScript frontend and a Python backend. Put more complicatedly, we use Node.js and Next.js to build the frontend capabilities and UI/UX elements. We then used Python and Google Collab along with the Dbias dataset and VaderSentiment to assemble the backend analysis. Finally, we used Flask and Node.js to connect the frontend server and backend python server together to make the website.
Challenges we ran into
There was a major issue with connecting the front and back end. We were heavily confused as to how exactly they were meant to communicate. We started with Firestore but slowly realized that we did not need a concurrent database. So, we moved from using a concurrent database to hosting our own webserver in Python. But this ran into more issues of missing imports and deprecated developer tools. Finally, we decided to host everything on Google Collab, which solved our issues. Another challenge was trying to work the API, we had to get one of the mentors to help quite heavily, because we id not really have a basis for it. The last major challenge we had was UI/UX design. We lack any design majors, so it was difficult to place everything correctly and figure out the exact sizing.
Accomplishments that we're proud of
We are proud that we have made a working website that connects to a hosted web server. We are very happy that we got Flask to work as an intermediary for our webservers as well. We are excited to have found applicable machine learning models. The final thing we are proud of is the UI/UX elements. We feel like we got the design across, and it looks visually clear and appealing.
What we learned
We learned that Webservers are very hard to make, but they are very rewarding. Webserver construction have very fun complexities, and although frustrating are great experiences to learn. Some specific skills we learned were how to make and use an API, how to connect a website with a webserver, and how to design specific UI elements like charts.
What's next for Gals Press
We want to deploy the website, add additional features, revise the article using open AI, have a user database to allow articles to be saved.
Log in or sign up for Devpost to join the conversation.