A Job listing aggregator that uses machine learning to provide Netflix-style rating estimation and similarity queries.
Want to try it out?
(Log in with guest/guest if you'd rather not create an account.)
JobSorter is built with the FAME stack (Flask, Angular, Mongo, Express).
The main components are:
- ai/aiserver.js:
Flask server that provides a JSON api to the other components. (Because python has the most machine learning support.) It processes the data in our Mongo database to generate similarity queries and numeric ratings. Uses GenSim and scikit-learn to do the heavy algorithmic lifting.
- scraper/getToday.js:
Script for grabbing job listings from Indeed.
- seeker/seeker.js
Interface for job-seekers. Built with AngularJS, Express.js, and a little Bootstrap for basic styling.