Recommender System Server: Extraction, Training, Deployment, and Prediction

This server extracts data, trains and deploys a model, and makes predictions using a matrix factorization recommender system.

ARCHITECTURE SUMMARY

• Data from MongoDB is preprocessed and converted into an interaction matrix, which is then stored in a BigQuery Datawarehouse.
• The interaction matrix is used to train a MATRIX FACTORIZATION model using the Stochastic Gradient Descent (SGD) algorithm.
• The trained model is then transferred to the Vertex AI model registry.
• From the model registry, the model is deployed to a Vertex AI prediction endpoint.
• The prediction endpoint is then used to make recommendations based on the interaction matrix.

Setup and Installation

Prerequisites

• Python >= 3.8
• Google Cloud SDK
• MongoDB
• BigQuery

Google Cloud SDK Authentication

• Local Server

1. Complete the SDK installation and setup
2. Run: gcloud auth application-default login

• Remote Server

1. Run: gcloud auth activate-service-account --key-file=path/to/the/service-account-key.json
2. Run: export GOOGLE_APPLICATION_CREDENTIALS="path/to/the/service-account-key.json"

Note: Place the service-account-key.json in the root directory. It will be ignored by git.

Environment Setup

1. Copy the .env.example file to .env and fill in the required values.
2. Create a Python virtual environment: python3 -m venv env
3. Activate the environment: source env/bin/activate
4. Install dependencies: pip install -r requirements.txt

Note: After installing any new dependencies, update the requirements.txt file: pip freeze > requirements.txt

Running the Application

To run the application, execute: python3 main.py

Contributing

When contributing to this project, please adhere to the following guidelines:

1. Fork the repository and create a new branch for your feature or bug fix.
2. Write clear, commented code and update documentation as necessary.
3. Submit a pull request with a detailed description of your changes.