Inspiration
Ever had a restaurant that just didn't match your vibe? Well we've had quite a few and decided to make a web app that matches your personality with certain aspects of a restaurant.
What it does
Our project takes in the location of the user, and has the user answer a variety of questions so we can match their personality to a restaurant! Restaurants within a certain radius of the user are collected and a variety of data is taken from them (mostly description and reviews) and an AI sorts through this to match certain personality traits to that restaurant.
How we built it
To collect the location data, we used JavaScript, and ported it over to a Python web scraper script to collect the restaurant data and put it into a JSON file. Next, we were planning on using Neurelo to act as the intermediary to get and send data to MongoDB. Lastly, we'd have Vectara add the final touches by actually matching the restaurants to the personalities, and output the top few restaurants for the user.
Challenges we ran into
Our group is fairly new to web development, and a lot of the time we spent on our project was dedicated to simply learning the basics as opposed to implementing our project's features. Our project was initially made with a mix of HTML, CSS, Javascript and Python, but we then spent a considerable amount of time porting it over to React. We likely could have gotten much further in our project with more prior experience in React/web development.
Accomplishments that we're proud of
We pushed our limits, and now that we've gone out of our way to learn these languages, we'll be able to use the lessons learned from this hackathon to better our future projects!
What we learned
Our group has gained some basic knowledge in React, Javascript, css, html, and web development overall. This should allow us to build web programs much faster in the future. We also dipped our toes into learning about some specific tools and services (like Neurelo and Vectara). It is unclear if we will use these services in the future, but we still had the opportunity to learn about and attempt to implement a new service.
What's next for RestaurantMBTI
We didn't actually get to get RestaurantMBTI to a fully-functional state so this would a first and foremost task to tackle. More features we could add would be letting users add their own restaurants into the database as web scraping can tend to leave out some of the newer restaurants.
Log in or sign up for Devpost to join the conversation.