UTE Learning Hub

Language / Ngôn ngữ: English | Tiếng Việt

A smart learning-material sharing platform for students, powered by AI to recommend relevant documents and study groups.

---

Key Features

For Students

• Microsoft Sign-in - Authenticate with the UTE student email account (@ute.udn.vn)
• Document management - Upload, share and search documents by subject, faculty and major
• Study groups - Create and join topic-based study groups
• Realtime chat - Message within groups with file and image support
• Smart recommendations (AI) - Suggest study groups and documents based on interests and behavior
• Personal library - Save favorite documents to read later
• Document reviews - Review and rate document quality
• Content reporting - Report documents/groups that violate the rules

For Admins

• Account management - View, assign roles and lock user accounts
• Document moderation - Approve or reject uploaded documents
• Report handling - Resolve violation reports from users
• Category management - CRUD for faculties, majors, subjects, document types and tags
• Analytics - Dashboard with charts analyzing system activity

---

Tech Stack

Layer	Technology
Frontend	Next.js 16, React 19, TypeScript, TailwindCSS 4, Radix UI
Backend	.NET 9.0, Clean Architecture, MediatR (CQRS), Entity Framework Core
AI Service	Python, FastAPI, Sentence Transformers (all-MiniLM-L6-v2)
Database	SQL Server 2022
Realtime	SignalR (WebSocket)
Auth	Microsoft Identity (MSAL)
Infrastructure	Docker, Nginx (Reverse Proxy), Let's Encrypt SSL

---

Project Structure

ute-learning-hub/
├── frontend/          # Next.js 16 - User interface
├── backend/           # .NET 9.0 - RESTful API (Clean Architecture)
│   ├── UteLearningHub.Api/           # Controllers, Middleware
│   ├── UteLearningHub.Application/   # Use Cases, Commands, Queries
│   ├── UteLearningHub.Domain/        # Entities, Events, Interfaces
│   ├── UteLearningHub.Infrastructure/# External Services, Email, AI Client
│   └── UteLearningHub.Persistence/   # EF Core, Repositories, Migrations
├── ai/                # Python FastAPI - AI Recommendation Service
├── nginx/             # Nginx configuration
├── docker-compose.yml # Development environment
└── docker-compose.prod.yml # Production environment

---

Getting Started

Requirements

• Docker & Docker Compose
• Git

Run with Docker

# Clone the repository
git clone https://github.com/your-org/ute-learning-hub.git
cd ute-learning-hub

# Copy the environment file
cp .env.example .env
# Edit .env with the appropriate values

# Build and start all services
docker-compose up -d --build

# View logs
docker-compose logs -f

Access the Application

Service	URL	Description
Frontend	http://localhost	Web interface
API Docs	http://localhost/scalar	Swagger/Scalar documentation
Health Check	http://localhost/health	Status check

---

Local Development

Backend (.NET 9.0)

cd backend
dotnet restore
dotnet run --project UteLearningHub.Api
# → http://localhost:7080

Frontend (Next.js 16)

cd frontend
npm install
npm run dev
# → http://localhost:3000

AI Service (Python)

cd ai
pip install -r requirements.txt
python main.py
# → http://localhost:8000/docs

---

Docker Services

Infrastructure

Service	Port	Description
SQL Server	1433	Main database

Application

Service	Internal port	Description
Backend	7080	.NET API
Frontend	3000	Next.js SSR
AI	8000	Recommendation Engine
Nginx	80, 443	Reverse Proxy & SSL

Managing Services

# Start everything
docker-compose up -d --build

# Start individual services
docker-compose up -d backend
docker-compose up -d frontend
docker-compose up -d ai

# View logs for a specific service
docker-compose logs -f backend

# Stop everything
docker-compose down

# Stop and remove volumes (reset data)
docker-compose down -v

---

Nginx Routing

Path	Service	Description
/	Frontend	Next.js pages
/api/*	Backend	REST API
/hubs/*	Backend	SignalR WebSocket
/images/*	Backend	Static files (cached 30 days)
/scalar	Backend	API Documentation

---

AI Recommendation System

The system uses Sentence Transformers (model all-MiniLM-L6-v2) to generate 384-dimensional vector embeddings for:

• User Vector: Based on the subjects and tags a user is interested in (from reading, upload and review history)
• Conversation Vector: Based on the group name, related subjects and tags

API Endpoints

# Create a user vector
POST /vector/user
{
  "subjects": ["Python Programming", "Machine Learning"],
  "subjectWeights": [10, 5],
  "tags": ["AI", "Data Science"],
  "tagWeights": [12, 8]
}

# Recommend study groups
POST /recommend
{
  "UserVector": [...],
  "ConversationVectors": [{"id": "xxx", "vector": [...]}],
  "TopK": 10,
  "MinSimilarity": 0.3
}

---

Database Schema

Main entities:

• AppUser - User information (linked with Microsoft Identity)
• Document - A document with multiple DocumentFiles
• DocumentFile - PDF/DOCX/PPTX file with metadata
• Conversation - Study group
• Message - Messages within a group (supports attachments)
• Subject, Faculty, Major, Type, Tag - Classification categories
• ProfileVector, ConversationVector - Vector embeddings for AI

---

Environment Variables

# Database
DB_SA_PASSWORD=YourStrong@Passw0rd
DB_NAME=ute-learning-hub

# Microsoft OAuth
MICROSOFT_CLIENT_ID=your-client-id
MICROSOFT_TENANT_ID=your-tenant-id

# Public URL (production)
PUBLIC_URL=https://your-domain.com