Version 1.0

collaborative-filtering-recommender

A collaborative filtering based recommender for recommending posts to the visitors on the page based on the data that can fetched using facebook APIs. User based collaborative filtering is used.

similarity measures

• pearson correlation similarity measure gave the best results so it is used.

Dependencies:

• Node Modules : express, body-parser, fb, graphfb
• Python Dependencies : sklearn, pandas, tensorflow, numpy, matplotlib

How to run:

1. Install the dependencies
2. Start the node server
3. Goto localhost for data fetching
4. To see a live dem of the project https://www.facebook.com/Python-Ladders-383643582063588/