W37L ML Service API

Overview

The W37L ML Service API is designed to provide machine learning functionalities for a decentralized social media platform. This service handles tasks such as profanity detection, hashtag generation, and recommendation systems, using a FastAPI framework with a strong emphasis on modularity and maintainability through dependency injection.

Prerequisites

To run this ML Service, you will need:

• Python 3.8 or higher
• FastAPI
• Uvicorn (ASGI server)
• Dependencies as listed in the requirements.txt file

Configuration

Before running the service, ensure your environment is set up with the necessary configurations:

export MLSERVICE_STORAGE_DIR=<path_to_storage_directory>
export ML_Service_LOG_LEVEL=INFO

Installation

Clone the repository and navigate to the project directory:

git clone https://github.com/W37L/W37L-ML-Service.git
cd W37L-ML-Service

Install the required Python packages:

pip install -r requirements.txt

Running the API

Start the ML service using Uvicorn:

uvicorn app:app --host 0.0.0.0 --port 8000 --reload

This command runs the service on localhost at port 8000 with live reloading enabled.

API Endpoints

The ML Service API provides several endpoints for machine learning tasks:

Profanity Check

• POST /MLService/checkProfanity Check text for profanity and return results.

Hashtag Generation

• GET /MLService/getHashtag Generate hashtags based on the provided text.

Trending Topics

• GET /MLService/getTrendingTopics Fetch current trending topics.

User Recommendations

• GET /MLService/recommendationUser Get user recommendations based on preferences.

Post Recommendations

• GET /MLService/recommendationPost Get post recommendations tailored to user interests.

CORS Middleware

The API is configured with CORS middleware to allow requests from all origins, which is suitable for development. For production, it is recommended to restrict origins to secure the API.

Deployment

For production environments, ensure that the MLSERVICE_STORAGE_DIR and logging levels are configured appropriately. It is also recommended to run the service behind a secure reverse proxy like Nginx.

Testing

You can test the API by accessing the live instance at http://ml.w37l.com:8000. This instance runs in a Docker container and includes Swagger UI for an interactive API documentation and testing experience. Use Swagger to view detailed information about each endpoint, including parameters and required data formats, or use Postman to send requests and analyze responses.

To test using Swagger:

1. Open a web browser and navigate to http://ml.w37l.com:8000/docs.
2. Explore the available API endpoints.
3. Execute API calls directly from the Swagger UI by entering required parameters and executing the request.

For Postman:

1. Open Postman and configure the request URL to http://ml.w37l.com:8000/<endpoint>.
2. Set the appropriate HTTP method and headers.
3. Send the request and view the response directly in Postman.

Contributing

Contributions to the ML Service API are welcome! Please fork the repository, create a new branch for your features, and submit a pull request.