wolverine-events

A simple content-based filtering recommender system for University of Michigan events. Write-up here.

In development. Hosted (sometimes) at mywolverine.events.

How to run

1. Create .env file with

GOOGLE_CLIENT_ID=[The app's client ID]
SESSION_SECRET=[key used to sign sessions with. For example, generated w/openssl rand -hex 32]
NTFY_CHANNEL=[ntfy.sh topic for maintenance notifications]

Note: last line must be empty!

2. Create email_params.json file with

{
    "token":[token used in script],
    "url":[appscript url that sends email],
    "subject":"Your recommendations are ready",
    "recipient":"[email protected]"
}

3. Create email.txt with weekly message. For example:

Hi,

Your weekly event recommendations are ready at https://mywolverine.events

Best,
Emilio

You received this email because you signed up to the site (https://mywolverine.events) and joined the MCommunity group https://mcommunity.umich.edu/group/Event%20Recommentations. To stop receiving these emails follow the group link and press "Leave Group".

4. Install dependencies w/pip install -r requirements.txt and run sudo sh serve.sh

TODOs

• Skip events user has already rated
• diversify feed (sample using predicted rating?)