Inspiration
We're tired of bad restaurant recommendations. We therefore wanted to create a tool that allows us to crowd-source an accurate consensus that is more impartial than a 5 star scale.
What it does
We built an API to query sentiments regarding specific aspects of a restaurant that matter to the user. this could be the food quality, price, service etc. The idea is that a user should be able to choose restaurants based on what matters to them and not some other person's criteria.
How we built it
A Flask app with 2 endpoints.
Challenges we ran into
Difficulty in scraping in a sustainable way.
Accomplishments that we're proud of
Creating an aspect-based sentiment analysis tool using deep learning trained tools available in the market.
What we learned
Heuristics are useful but not foolproof. In the field of NLP, we still need linguistics majors to help us create useable models of language.
What's next for Foodify
That's for the HNR team to decide :)
Log in or sign up for Devpost to join the conversation.