Game Sage

A machine learning, video game recommendation system.

Overview

Game Sage works by reading in data from a PostgreSQL table, clustering the games using k-means and then writing the results out to Elasticsearch.

Running

If you want to run Game Sage, you will have to setup a PostgreSQL instance, run the table.sql script and then load game data from the GiantBomb API (https://www.giantbomb.com/api/games).

Once you have the data, start up an Elasticsearch instance

sudo ./elasticsearch.sh

and then run main.py, supplying a PostgresSQL connection string

python main.py "host='localhost' dbname='db' user='user' password='password'"

Results

You can view the results using the Game Sage front-end.