Pangea Recommender System

Recommender System - Marketplace Matching Script for Pangea.App

• Recommender_lib.py - Python Script containing Word2Vec model and recommender system
• app.py - flask application for deploying the recommender system; Flask App for Pangea.App
• requirements.txt - requirements needed to run the model
• Pangea_RS_Cleaned.ipnyb - Jupyter Notebook that outlines word2vec model and data output line by line
• App_Notebook_Viz.ipynb - Jupyter notebook with more detailed explanation of each part of the code.

Notes for running flask app:

• remember to check both the client and server side
• use the curl command to call on the flask app: curl localhost:5000/pangeaapp -d '{"title": "Teach me how to cook!"}' -H 'Content-Type: application/json'

Transfer Learning:

• Download the Google Vectors Bin file here https://drive.google.com/uc?id=0B7XkCwpI5KDYNlNUTTlSS21pQmM&export=download