Inspiration
Tinder with a twist.
What it does
Users create an account and upload a picture to our database. Microsoft API analyses this picture to give a list of emotions displayed. These emotions are compared to other user pictures in 8-dimensional vector space: the closest 3 points are automatically connected as matches -- the furthest point away is also connected as a wildcard match. Conversations are analysed on the fly with the Microsoft sentiment API to determine how likely a date is.
How we built it
Ionic framework for UI and AJAX, Python and flask for server and logic, firebase for database.
Challenges we ran into
Novice programmers in making mobile apps. Encountered some trouble in organising the database and integrating our functioning algorithms into a sleek UI
What we learned
How to handle API requests/responses, using firebase as a real-time database
What's next for Ratio
Exploration of accuracy of matching algorithm and possible UI improvements and bug fixes
Log in or sign up for Devpost to join the conversation.