SportsHack

During SportsHack, built a web application that allows fans to immerse themselves into a) the world of their favorite player if they are familiar with the CFL or b) have the choice to be matched with a player based of the commonality of personal information if they are new to the CFL. Used cosine-similarity (Machine Learning) to produce a list of top 5 most similar players to the fan based on given sign-up input data (alma matter, hometown, etc).

Languages/Frameworks: Front-end: HTML, CSS, JavaScript, jQuery Back-end: Parse API, Node.js, Python, Flask, Heroku (RESTful API) Data Analysis: Cosine Similarity Libraries: Various CFL .csv's and GeoPy


Created new repo of branch worked on during SportsHack 2015.

Note: This is a beta version, calls to Social Media APIs (Youtube, Twitter, Instagram etc) for dynamically generating content are in development. Once this is encorporated, Parse will be able to store content from each CFL user profile.

Alt text

Share this project:

Updates