Recko

Calculates distance between articles and recommend new articles to read

Teach some article url's to recko
Calculates the distance between teached articles according to keywords extracted
Then get closest neighbors of given articles
brings recommendations according to user's liked articles

Example

Setup

copy the config.py.copy as config.py
change inside according to your config details
run app.py for flask web server
run teacher.py for teaching articles
run calculate.py for distance calculating worker

Usage

For teaching urls GET /teach?url=<path-to-url>
For getting one of the closest neighbors of given url GET /neighbor?url=<url-id>
Authenticate a new client POST /authenticate {} returns {token: <access token>}
Get a suggestion GET /next?token=<access_token> returns a response according to your likes
Like an article POST /like?token=<access_token> {url: <article_id>}

Needs

numpy
scipy
mongodb
redis
embedly account for extracting kw's

Contributors