Inspiration

I wanted to address the COVID-19, Best Financial Hack, and Nittany AI challenge. During the COVID-19 pandemic, many people found it difficult to find masks, hand sanitizers, paper towels, etc to stay protected. So I built a web app to help them out!

What it does

My web app has two tools that a user can use, an item finder program and a COVID-19 Symptoms checker. The Item finder lets a user choose an item that they are looking for and enter their city name, then the program will use the SerpAPI to locate stores in that city that have those items in stock, and output a list of these stores. The COVID-19 Symptoms Checker allows the user to fill out a questionnaire, then uses a Nearest Neighbor Classifier to predict whether the user is at risk of the virus.

How I built it

I built the web app using the Streamlit library while also using an MVC design pattern. Although this library is used for data science I manipulated a few components to make it seem more of a regular website. The Item Finder is built around the SerpAPI which allows me to web scrape the Google Shopping site and get nearby stores that sell the item I am looking for. I then stored each item in an ItemData class to make it easier to reference the data when displaying onto the web app. The Symptoms Checker uses a Nearest Neighbor Classifier from Sklearn that I trained with COVID-19 patient data from Kaggle. Using a few optimization methods, I was able to improve the model to have a 97% accuracy.

Challenges I ran into

One challenge that I ran into was finding a way to search through many stores to find whether they have the item the user is looking for. Luckily I came upon the SerpAPI which does most of the work for me, all I needed to do is specify it to look through the Google Shopping site for items and return the stores which have that item available.

Accomplishments that I'm proud of

Other than the Covid-19 classifier, is having emojis appear on the app.

What I learned

How to build a web app using Streamlit and learning new optimization methods to improve machine learning classifiers.

What's next for The Pandemic Assistant

Adding a network system for users to chat with other users.

Built With

Share this project:

Updates