Emati is a recommender system for biomedical literature. The project is built on Django. It uses django-allauth for user management and Elasticsearch to provide a search for papers.

Project Structure

The project is divided into the following apps:

• website: The main app.
• machinelearning: A custom python module that handles everything related to training and using machine learning classifiers.
• fetching: A custom python module that handles everything related to getting articles from various sources and storing them in our database.
• dashboard: A separate website that displays statistics about the project.

+ emati
  + docs      (documentation)
  + emati     (main project)
    + apps
      + dashboard        (Django app)
      + fetching         (custom python module)
      + machinelearning  (custom python module)
      + website          (Django app)
    + settings
  + media       (user uploads, classifiers, ...)
  + static      (js, css, images)
  + templates   (html templates)
  + tests       (unit tests)

Tests

To run the tests against the current environment:

python manage.py test