Self-Learning Classifier

This is a full-stack web application that allows users to classify objects into two categories based on their textual features. The system learns and improves in real-time from user feedback.

The project consists of a Go backend providing a REST API and a React (TypeScript) frontend. The entire stack, including a MySQL database, is containerized with Docker for easy setup and deployment.

💡 Key Features

• 🧠 Real-time Learning: The system updates its knowledge base after every user-confirmed classification.
• ↔️ Feature Separation: Automatically identifies common properties shared between both classes, unique properties for each, and irrelevant ("none") properties.
• ⚙️ Flexible Management: The UI allows for adding, removing, moving, and renaming properties and classes on the fly.
• 🚀 Modern Stack: Built with Go for the backend, React/Vite/TS for the frontend, and MySQL for persistence.
• 🐳 Fully Containerized: The entire project runs with a single docker-compose up command.

🛠️ Tech Stack

• Backend: Go, net/http, go-sql-driver/mysql
• Frontend: React, TypeScript, Vite, Nginx (for serving static files)
• Database: MySQL 8.0
• DevOps: Docker, Docker Compose

🚀 Quick Start with Docker

This is the recommended way to run the project for demonstration or development. Ensure you have Docker and Docker Compose installed.

1. Clone the repository:

   git clone https://github.com/AntonHritsai/self-learning-classifier.git
   cd self-learning-classifier

2. Set up environment variables: Create a .env file by copying the example. The default values are already configured for local Docker development.

   cp .env.example .env

3. Launch the project: This command will build the images, start all containers, and set up the network.

   docker-compose up --build

4. Access the application:

   • Frontend: http://localhost:3000
   • Backend API: http://localhost:8080

🛠️ Local Development (Without Docker)

If you prefer to run the services natively, you will need:

• Go 1.22+
• Node.js 20+
• A running MySQL 8.0 instance

Backend

1. Navigate to the Backend directory:

   cd Backend

2. Install dependencies:

   go mod tidy

3. Set the required environment variables to connect to your MySQL instance. For example:

   export DB_HOST=127.0.0.1
   export DB_PORT=3306
   export DB_USER=your_user
   export DB_PASSWORD=your_password
   export DB_NAME=slc

4. Run the server:

   go run ./cmd/api

   The backend will be available at http://localhost:8080.

Frontend

1. Navigate to the Frontend directory:

   cd Frontend/slc-frontend

2. Install dependencies:

   npm install

3. Create a .env.local file and specify the backend API URL:

   VITE_API_URL=http://localhost:8080
   

4. Run the development server:

   npm run dev

   The frontend will be available at http://localhost:5173 (or another port specified by Vite).

🧪 Testing

The Go backend includes a suite of unit tests for handlers, services, and core logic. The tests use in-memory mocks for the repository to ensure isolation and speed.

To run the tests, navigate to the Backend directory and execute:

cd Backend
go test -v ./...

📁 Project Structure

.
├── Backend/                # Go Backend
│   ├── cmd/api/            # Application entry point
│   ├── internal/           # All business logic
│   └── ...
├── Frontend/
│   └── slc-frontend/       # React Application
│       ├── src/            # Frontend source code
│       └── ...
├── docs/                   # Documentation and assets
├── .env.example            # Example environment file
└── docker-compose.yml      # Orchestration file for all services

⚙️ Configuration

The application is configured using environment variables. See .env.example for a complete list. Key variables include:

Variable	Default	Description
\DB_HOST	db	Hostname of the MySQL database service.
\DB_USER	slc	MySQL user.
\MYSQL_PASSWORD	slcpass	Password for the MySQL user.
\DB_NAME	self-learning-classifier	Name of the database.
\BACKEND_PORT	8080	Port on which the Go backend listens.
\FRONTEND_PORT	3000	Port on which the Nginx frontend is exposed.
\VITE_API_URL	http://localhost:8080\	URL of the backend API for the frontend to use.

📚 API Endpoints

Base URL: /api/v1

Method	Path	Description
\POST	/init	Initializes the classes and their seed properties.
\POST	/reset	Resets the state for the current user.
\POST	/classify	Classifies a given set of properties.
\POST	/feedback	Provides feedback to train the model.
\GET	/state	Retrieves the current state of the classifier.
\POST	/prop/add	Adds a new property to a specific area.
\POST	/prop/remove	Removes a property from a specific area.
\POST	/prop/move	Moves a property between areas.
\POST	/prop/rename	Renames a property within an area or globally.
\POST	/classes/rename	Renames a class.
\GET	/status	Health check endpoint.