zcdlab / docker-predictionio_with_universal_recommender

Run PredictionIO with Universal Recommender inside Docker

docker-predictionio

Run PredictionIO inside Docker

1. Run build to build the image
2. Run shell to start the container
3. Once inside the container, run runsvdir-start& to start everything
4. The Dashboard is available on port 9000

Run Prediction IO Recommendation Example

1. Go to quickstartapp directory cd /quickstartapp
2. Build and Train Engine ./run.sh
3. Deploy Engine cd MyRecommendation && pio deploy --ip 0.0.0.0&
4. Your Engine will now listen on port 8000

Run ActionML Univerisal Recommender Example

Train and Deploy the Model

1. cd /other-engines/ur/
2. ./examples/integration-test
3. pio deploy

Query Data

• ./examples/multi-query-handmade.sh

curl -H "Content-Type: application/json" -d '
{
    "user": "u1"
}' http://localhost:8000/queries.json