/PV254-recommender-system

Primary LanguagePython

PV254-recommender-system

Heroku

We had some problems with a image uploading to Heroku so during evaluation of a project it is hosted here.

Cloud Vision API dependencies

  • env variable GOOGLE_APPLICATION_CREDENTIALS needs to be set to the path where service account key is stored
  • it is sufficient to place access.json inside .secrets folder

Setup

  1. Install virtualenv via pip install virtualenv (alternatively pip3 install virtualenv).
  2. Create new virtual environment virtualenv venv (alternatively virtualenv -p python3 venv).
  3. Activate virtual environment source venv/bin/activate (Windows: venv\Scripts\activate).
  4. Install all requirements via pip install -r requirements.txt
  5. Leaving virtual environment: deactivate.

Example usage

  1. Export PYTHONPATH to root dir export PYTHONPATH=`pwd` .
  2. Download random images python scripts/download_random_photos.py.
  3. Download Instagram profile python scripts/download_ig_profile.py --profile jpancik.
  4. Annotate random images python scripts/annotate_images.py --input resources/random-images/.
  5. Annotate IG profile images python scripts/annotate_images.py --input profiles/jpancik/.
  6. Rank images python scripts/rank_images.py --library profiles/jpancik/ --candidates random-images/.
  7. Open HTML with results in resources/random-images/recommendations.html.
  8. $$$ Profit $$$.

Django frontend

  1. Perform all steps in Setup
  2. Run server:
    • dev python manage.py runserver is listening on 127.0.0.1:8000
    • prod gunicorn --timeout 300 frontend.wsgi

Export profiles to CSV

  1. python scripts/export_label_count_from_profiles_to_csv.py --dbfile <PATH_TO_DB_FILE> --output <PATH_TO_EXPORT_CSV> --min_score <OPTIONAL_MIN_SCORE>

How to use "Galajdator"

  1. Download model from https://www.dropbox.com/s/buix0deqlks4312/lexvec.commoncrawl.ngramsubwords.300d.W.pos.bin.gz?dl=1
  2. Unzip it to /lib/lexvec_model/lexvec_model.bin