📚 TeachAI - AI-Powered Lesson Plan Generator

A full-stack AI-powered lesson plan generation application built with Next.js and Node.js, featuring an integrated RAG (Retrieval-Augmented Generation) system for more accurate educational content generation.

🚀 Quick Start

Prerequisites

Node.js >= 18.17.0
pnpm >= 8.0.0
Python >= 3.8 (for ChromaDB)

📦 Package Manager Installation

This project uses pnpm as the package manager. If you haven't installed pnpm yet, please install it first:

# Install pnpm globally
npm install -g pnpm

# Or use Corepack (Node.js 16.10+)
corepack enable
corepack prepare pnpm@latest --activate

🔧 Project Installation & Setup

Clone the repository

git clone https://github.com/your-username/teachai.git
cd teachai

Install all dependencies

# Install dependencies for root, web, and server
pnpm run install:all

# Or install separately
pnpm install # Root dependencies
cd web && pnpm install # Frontend dependencies
cd ../server && pnpm install # Backend dependencies

Install ChromaDB (RAG System)

# Install ChromaDB
pip install chromadb

# Or use conda
conda install -c conda-forge chromadb

Configure environment variables

# Copy environment template
cp server/.env.example server/.env

# Edit environment variables, set AI API key
# DASHSCOPE_API_KEY=your_api_key_here

🧠 RAG System Setup

This project integrates an advanced RAG (Retrieval-Augmented Generation) system using ChromaDB as the vector database, providing intelligent content generation based on educational materials.

Starting the RAG System

Start ChromaDB service

# Start ChromaDB vector database
pnpm run chroma:start

Load educational materials data

# Load educational materials from enhanced chunks to vector database
pnpm run rag:load

Check RAG system status

# Check database status and data statistics
pnpm run rag:status

One-click RAG System Setup

# Automatically start ChromaDB and load data
pnpm run setup:rag

RAG System Management

# Start ChromaDB service
pnpm run chroma:start

# Stop ChromaDB service
pnpm run chroma:stop

# Check system status
pnpm run rag:status

# Reload data
pnpm run rag:load

# Monitor loading progress (live updates)
pnpm run rag:progress:watch

# Verify data integrity and performance
pnpm run rag:verify

# Run comprehensive RAG system tests
pnpm run rag:test

🏃‍♂️ Running the Project

Development Mode

Standard development mode (without RAG)

# Start both frontend and backend development servers
pnpm dev

Full development mode (with RAG)

# Start ChromaDB, frontend, and backend servers simultaneously
pnpm run dev:full

Starting Services Separately

# Start ChromaDB service
pnpm run chroma:start

# Start backend server
pnpm run dev:server

# Start frontend server
pnpm run dev:web

Access URLs:

Frontend application: http://localhost:3000
Backend API: http://localhost:3001
ChromaDB: http://localhost:8000

📚 Educational Materials Data

Data Structure

The project's server/rag_data/chunks/ folder contains enhanced educational materials data in JSON format:

server/rag_data/chunks/
├── enhanced_chunk_0001.json
├── enhanced_chunk_0002.json
├── enhanced_chunk_0003.json
└── ... (95,360+ educational chunks)

Data Format

Each JSON file contains the following structure:

{
  "content": "Educational content text",
  "metadata": {
    "source": "filename",
    "page_number": 1,
    "chunk_index": 0,
    "qualityScore": 0.85,
    "ocrConfidence": 0.92,
    "semanticFeatures": ["definitions", "examples"]
  }
}

Supported Textbook Versions

Mathematics: People's Education Press, Beijing Normal University, Hunan Education, East China Normal University
Chinese: People's Education Press, Jiangsu Education, Beijing Normal University
English: People's Education Press, Foreign Language Teaching and Research Press, Yilin Press
Physics: People's Education Press, Beijing Normal University, Shanghai Science
Chemistry: People's Education Press, Shandong Science and Technology
Biology: People's Education Press, Jiangsu Education
Others: Music, Art, Science, etc.

Enhanced Data Features (Version 2.0)

OCR Error Correction: Improved text accuracy for Chinese educational content
Duplicate Detection: Advanced algorithms to identify and merge similar content
Quality Scoring: Each chunk includes reliability metrics (0.3-1.0 scale)
Content Classification: Semantic features like formulas, experiments, definitions
Smart Chunking: Optimized text segmentation for better retrieval
95,360+ Enhanced Chunks: Comprehensive K-12 curriculum coverage

Build Project

# Build entire project
pnpm build

# Build separately
pnpm run build:web     # Build frontend
pnpm run build:server  # Build backend

🎯 Technical Features

RAG (Retrieval-Augmented Generation) System

This project integrates an advanced RAG system that retrieves relevant educational materials through a vector database, significantly improving the quality and accuracy of AI-generated content.

Core Features

Intelligent Retrieval: Semantic similarity-based retrieval of relevant educational content
Multi-version Textbook Support: Covers mainstream textbooks like People's Education Press, Beijing Normal University, Jiangsu Education, etc.
Grade Adaptation: Automatically matches grade-relevant educational materials
Subject Specialization: Provides specialized educational content for different subjects
Real-time Enhancement: Real-time retrieval and integration of relevant educational resources during generation

Data Statistics

Educational Materials: 95,360+ enhanced educational content chunks
Grade Coverage: Elementary Grade 1 to High School Grade 12
Supported Subjects: Mathematics, Chinese, English, Physics, Chemistry, Biology, History, Geography, Politics, Music, Art, Science
Textbook Versions: 20+ mainstream textbook versions

Technical Architecture

User Request → AI Service → Vector Retrieval → Content Fusion → Enhanced Generation → Return Result
                ↓
        ChromaDB Vector Database
        (95,360+ educational materials)

AI Generation Capabilities

Lesson Plan Generation: Supports complete instructional design including teaching objectives, key points, and teaching process
Exercise Generation: Intelligent question generation supporting multiple question types and difficulty levels
Content Analysis: Intelligent analysis of educational content, extracting core concepts
Multi-format Output: Supports text, mind maps, flowcharts, timelines, and other display formats

User Experience

Streaming Output: Real-time display of generation process, enhancing user experience
Personalized Settings: Customize generated content based on user preferences
Content Management: Complete content bookmarking, export, and deletion functionality
Multi-format Export: Supports PDF, Word, image, and other export formats

📖 Usage Guide

Basic Usage Flow

Start Services

# Start complete services (recommended)
pnpm run dev:full

Generate Lesson Plans
- Visit http://localhost:3000
- Select subject, grade, and topic
- Click generate, AI will provide professional lesson plans combined with RAG system
Manage Content
- View generation history
- Bookmark quality content
- Export as PDF/Word

Advanced Features

RAG System Management: Use pnpm run rag:status to check database status
Data Updates: Use pnpm run rag:load to reload educational materials
System Monitoring: View generation logs and performance metrics

Production Environment Startup

# Start production environment
pnpm start

# Start separately
pnpm run start:web     # Start frontend production server
pnpm run start:server  # Start backend production server

📁 Project Structure

teachai/
├── web/              # Next.js frontend application
│   ├── src/
│   │   ├── app/      # App Router pages
│   │   └── components/ # React components
│   ├── package.json
│   └── pnpm-lock.yaml
├── server/           # Node.js backend service
│   ├── models/       # Data models
│   ├── services/     # Business services
│   ├── middleware/   # Middleware
│   ├── config/       # Configuration files
│   ├── utils/        # Utility functions
│   ├── rag/          # RAG system implementation
│   ├── rag_data/     # Enhanced educational materials (95,360+ chunks)
│   ├── package.json
│   └── pnpm-lock.yaml
├── package.json      # Root configuration (workspace)
├── pnpm-lock.yaml    # Lock file
└── .gitignore        # Git ignore file

🛠️ Development Tools

Available Scripts

pnpm dev              # Start frontend and backend in development mode
pnpm build            # Build production version
pnpm start            # Start production servers
pnpm lint             # Code style check
pnpm test             # Run tests
pnpm clean            # Clean node_modules and build files
pnpm format           # Code formatting

Git Ignore Rules

The project has a complete .gitignore file configured to automatically ignore:

**/node_modules/ - Dependency directories at all levels
**/.next/ - Next.js build output
**/*.log - Log files
.env* - Environment variable files
.DS_Store - macOS system files
IDE configuration files, etc.

🔑 Environment Configuration

Please create a .env file in the server/ directory:

# MongoDB connection
MONGODB_URI=your_mongodb_connection_string

# JWT secret
JWT_SECRET=your_jwt_secret

# Qwen API Key (Alibaba Cloud DashScope)
DASHSCOPE_API_KEY=your_dashscope_api_key

# Server port
PORT=3001

# AI Configuration
QWEN_MODEL=qwen-plus
AI_MAX_TOKENS=2000
AI_TEMPERATURE=0.7

# ChromaDB Configuration (Local)
CHROMADB_HOST=localhost
CHROMADB_PORT=8000

# ChromaDB Cloud Configuration (Production)
CHROMA_CLOUD_ENABLED=false
CHROMADB_API_KEY=your_chromadb_cloud_api_key
CHROMADB_TENANT=your_tenant_id
CHROMADB_DATABASE=teachai
CHROMADB_COLLECTION=teachai_main

📖 Technology Stack

Frontend (web/)

Next.js 15 - React full-stack framework
React 19 - UI library
TypeScript - Type safety
Tailwind CSS - Styling framework
Framer Motion - Animation library
Lucide React - Icon library

Backend (server/)

Node.js - Runtime environment
Express.js - Web framework
MongoDB - Database
Mongoose - ODM
JWT - Authentication
Winston - Logging
ChromaDB - Vector database for RAG

RAG System

ChromaDB - Vector database
OpenAI Embeddings - Text embeddings
Enhanced Educational Data - 95,360+ quality-scored chunks
Semantic Search - Intelligent content retrieval

🚀 Deployment

Local Development Deployment

pnpm build
pnpm start

ChromaDB Cloud Production Deployment

For production deployment with ChromaDB Cloud, see the comprehensive Production Deployment Guide.

Quick Start for ChromaDB Cloud

Set Environment Variables:

export CHROMA_CLOUD_ENABLED=true
export CHROMADB_API_KEY=your_chromadb_api_key
export CHROMADB_TENANT=your_tenant_id
export CHROMADB_DATABASE=teachai
export CHROMADB_COLLECTION=teachai_main

Test Configuration:

node test-env-config.js

Upload RAG Data to Cloud:

node server/rag/scripts/cloud-uploader.js

Deploy Application:

NODE_ENV=production pnpm build && pnpm start

🧪 Testing

Running Tests

# Run all tests
pnpm test

# Run tests with coverage
pnpm test:coverage

# Run server tests only
pnpm run test:server

# Run web tests only
pnpm run test:web

# Run single test file
cd web && pnpm test -- --testPathPattern=page.test.tsx
cd server && pnpm test -- --testPathPattern=ai-service.test.js

RAG System Testing

# Test RAG accuracy with quality scoring
pnpm run rag:test-accuracy

# Run comprehensive RAG system tests
pnpm run rag:test

# Verify loaded data integrity and performance
pnpm run rag:verify

🤝 Contributing

Welcome to submit Pull Requests and Issues!

Development Workflow

Fork the project
Create a feature branch
Commit your changes
Push to the branch
Create a Pull Request

📄 License

MIT License

📞 Contact Us

Project Link: GitHub Repository
Issue Reports: Issues

TeachAI - Let AI become your teaching assistant 🚀

中文版本 / Chinese Version

📚 TeachAI - 智能教案生成器

基于 Next.js 和 Node.js 的全栈 AI 教案生成应用，集成 RAG (检索增强生成) 系统，提供更准确的教学内容生成。

🚀 快速开始

环境要求

Node.js >= 18.17.0
pnpm >= 8.0.0
Python >= 3.8 (用于 ChromaDB)

📦 包管理器安装

项目使用 pnpm 作为包管理器。如果您还没有安装 pnpm，请先安装：

# 全局安装 pnpm
npm install -g pnpm

# 或者使用 Corepack (Node.js 16.10+)
corepack enable
corepack prepare pnpm@latest --activate

🔧 项目安装与运行

克隆项目

git clone https://github.com/your-username/teachai.git
cd teachai

安装所有依赖

# 安装根目录、web 和 server 的所有依赖
pnpm run install:all

# 或者分别安装
pnpm install # 根目录依赖
cd web && pnpm install # 前端依赖
cd ../server && pnpm install # 后端依赖

安装 ChromaDB (RAG 系统)

# 安装 ChromaDB
pip install chromadb

# 或者使用 conda
conda install -c conda-forge chromadb

配置环境变量

# 复制环境变量模板
cp server/.env.example server/.env

# 编辑环境变量，设置 AI API 密钥
# DASHSCOPE_API_KEY=your_api_key_here

🧠 RAG 系统设置

本项目集成了 RAG (检索增强生成) 系统，使用 ChromaDB 作为向量数据库，提供基于教学材料的智能内容生成。

启动 RAG 系统

启动 ChromaDB 服务

# 启动 ChromaDB 向量数据库
pnpm run chroma:start

加载教学材料数据

# 加载增强版教学材料到向量数据库
pnpm run rag:load

检查 RAG 系统状态

# 检查数据库状态和数据统计
pnpm run rag:status

一键设置 RAG 系统

# 自动启动 ChromaDB 并加载数据
pnpm run setup:rag

RAG 系统管理

# 启动 ChromaDB 服务
pnpm run chroma:start

# 停止 ChromaDB 服务
pnpm run chroma:stop

# 检查系统状态
pnpm run rag:status

# 重新加载数据
pnpm run rag:load

# 实时监控加载进度
pnpm run rag:progress:watch

# 验证数据完整性和性能
pnpm run rag:verify

# 运行综合 RAG 系统测试
pnpm run rag:test

🏃‍♂️ 运行项目

开发模式

标准开发模式 (不含 RAG)

# 同时启动前端和后端开发服务器
pnpm dev

完整开发模式 (含 RAG)

# 同时启动 ChromaDB、前端和后端服务器
pnpm run dev:full

分别启动各服务

# 启动 ChromaDB 服务
pnpm run chroma:start

# 启动后端服务器
pnpm run dev:server

# 启动前端服务器
pnpm run dev:web

访问地址：

前端应用：http://localhost:3000
后端 API：http://localhost:3001
ChromaDB：http://localhost:8000

📚 教学材料数据

数据结构

项目的 server/rag_data/chunks/ 文件夹包含了增强版教学材料数据，格式为 JSON 文件：

server/rag_data/chunks/
├── enhanced_chunk_0001.json
├── enhanced_chunk_0002.json
├── enhanced_chunk_0003.json
└── ... (95,360+ 教学内容片段)

数据格式

每个 JSON 文件包含以下结构：

{
  "content": "教学内容文本",
  "metadata": {
    "source": "文件名",
    "page_number": 1,
    "chunk_index": 0,
    "qualityScore": 0.85,
    "ocrConfidence": 0.92,
    "semanticFeatures": ["definitions", "examples"]
  }
}

支持的教材版本

数学: 人教版、北师大版、湘教版、华师大版
语文: 人教版、苏教版、北师大版
英语: 人教版、外研版、译林版
物理: 人教版、北师大版、沪科版
化学: 人教版、鲁科版
生物: 人教版、苏教版
其他: 音乐、美术、科学等

增强版数据特性 (版本 2.0)

OCR 错误纠正: 提高中文教育内容的文本准确性
重复检测: 高级算法识别和合并相似内容
质量评分: 每个片段包含可靠性指标 (0.3-1.0 评分)
内容分类: 语义特征如公式、实验、定义等
智能分块: 优化文本分割以提高检索效果
95,360+ 增强片段: 全面覆盖 K-12 课程内容

构建项目

# 构建整个项目
pnpm build

# 分别构建
pnpm run build:web     # 构建前端
pnpm run build:server  # 构建后端

🎯 技术特性

RAG (检索增强生成) 系统

本项目集成了先进的 RAG 系统，通过向量数据库检索相关教学材料，显著提升AI生成内容的质量和准确性。

核心特性

智能检索: 基于语义相似度检索相关教学内容
多版本教材支持: 覆盖人教版、北师大版、苏教版等主流教材
年级适配: 自动匹配年级相关的教学材料
学科专业化: 针对不同学科提供专业化的教学内容
实时增强: 生成过程中实时检索和融合相关教学资源

数据统计

教学材料数量: 95,360+ 个增强教学内容片段
覆盖年级: 小学一年级至高中三年级
支持学科: 数学、语文、英语、物理、化学、生物、历史、地理、政治、音乐、美术、科学
教材版本: 20+ 种主流教材版本

技术架构

用户请求 → AI服务 → 向量检索 → 内容融合 → 增强生成 → 返回结果
            ↓
      ChromaDB向量数据库
      (95,360+ 教学材料)

AI 生成能力

教案生成: 支持完整的教学设计，包含教学目标、重难点、教学过程
练习题生成: 智能出题，支持多种题型和难度级别
内容分析: 智能分析教学内容，提取核心概念
多格式输出: 支持文本、思维导图、流程图、时间线等多种展示格式

用户体验

流式输出: 实时显示生成过程，提升用户体验
个性化设置: 根据用户偏好定制生成内容
内容管理: 完整的内容收藏、导出、删除功能
多格式导出: 支持PDF、Word、图片等多种导出格式

📖 使用指南

基本使用流程

启动服务

# 启动完整服务（推荐）
pnpm run dev:full

生成教案
- 访问 http://localhost:3000
- 选择学科、年级、主题
- 点击生成，AI将结合RAG系统提供专业教案
管理内容
- 查看生成历史
- 收藏优质内容
- 导出为PDF/Word

高级功能

RAG系统管理: 使用 pnpm run rag:status 查看数据库状态
数据更新: 使用 pnpm run rag:load 重新加载教学材料
系统监控: 查看生成日志和性能指标

生产环境启动

# 启动生产环境
pnpm start

# 分别启动
pnpm run start:web     # 启动前端生产服务器
pnpm run start:server  # 启动后端生产服务器

📁 项目结构

teachai/
├── web/              # Next.js 前端应用
│   ├── src/
│   │   ├── app/      # App Router 页面
│   │   └── components/ # React 组件
│   ├── package.json
│   └── pnpm-lock.yaml
├── server/           # Node.js 后端服务
│   ├── models/       # 数据模型
│   ├── services/     # 业务服务
│   ├── middleware/   # 中间件
│   ├── config/       # 配置文件
│   ├── utils/        # 工具函数
│   ├── rag/          # RAG 系统实现
│   ├── rag_data/     # 增强版教学材料 (95,360+ 片段)
│   ├── package.json
│   └── pnpm-lock.yaml
├── package.json      # 根目录配置（工作空间）
├── pnpm-lock.yaml    # 锁定文件
└── .gitignore        # Git 忽略文件

🛠️ 开发工具

可用脚本

pnpm dev              # 开发模式启动前后端
pnpm build            # 构建生产版本
pnpm start            # 启动生产服务器
pnpm lint             # 代码风格检查
pnpm test             # 运行测试
pnpm clean            # 清理 node_modules 和构建文件
pnpm format           # 代码格式化

Git 忽略规则

项目已配置完整的 .gitignore 文件，自动忽略：

**/node_modules/ - 所有层级的依赖目录
**/.next/ - Next.js 构建输出
**/*.log - 日志文件
.env* - 环境变量文件
.DS_Store - macOS 系统文件
IDE 配置文件等

🔑 环境配置

请在 server/ 目录下创建 .env 文件：

# MongoDB 连接
MONGODB_URI=your_mongodb_connection_string

# JWT 密钥
JWT_SECRET=your_jwt_secret

# 通义千问 API Key (阿里云 DashScope)
DASHSCOPE_API_KEY=your_dashscope_api_key

# 服务器端口
PORT=3001

# AI 配置
QWEN_MODEL=qwen-plus
AI_MAX_TOKENS=2000
AI_TEMPERATURE=0.7

# ChromaDB 配置 (本地)
CHROMADB_HOST=localhost
CHROMADB_PORT=8000

# ChromaDB 云端配置 (生产环境)
CHROMA_CLOUD_ENABLED=false
CHROMADB_API_KEY=your_chromadb_cloud_api_key
CHROMADB_TENANT=your_tenant_id
CHROMADB_DATABASE=teachai
CHROMADB_COLLECTION=teachai_main

📖 技术栈

前端 (web/)

Next.js 15 - React 全栈框架
React 19 - UI 库
TypeScript - 类型安全
Tailwind CSS - 样式框架
Framer Motion - 动画库
Lucide React - 图标库

后端 (server/)

Node.js - 运行时环境
Express.js - Web 框架
MongoDB - 数据库
Mongoose - ODM
JWT - 身份验证
Winston - 日志记录
ChromaDB - RAG 向量数据库

RAG 系统

ChromaDB - 向量数据库
OpenAI Embeddings - 文本嵌入
增强版教育数据 - 95,360+ 质量评分片段
语义搜索 - 智能内容检索

🚀 部署

本地开发部署

pnpm build
pnpm start

ChromaDB 云端生产部署

有关 ChromaDB 云端生产部署的详细信息，请参阅综合生产部署指南。

ChromaDB 云端快速开始

设置环境变量:

export CHROMA_CLOUD_ENABLED=true
export CHROMADB_API_KEY=your_chromadb_api_key
export CHROMADB_TENANT=your_tenant_id
export CHROMADB_DATABASE=teachai
export CHROMADB_COLLECTION=teachai_main

测试配置:

node test-env-config.js

上传 RAG 数据到云端:

node server/rag/scripts/cloud-uploader.js

部署应用:

NODE_ENV=production pnpm build && pnpm start

🧪 测试

运行测试

# 运行所有测试
pnpm test

# 运行测试并生成覆盖率报告
pnpm test:coverage

# 仅运行服务器测试
pnpm run test:server

# 仅运行前端测试
pnpm run test:web

# 运行单个测试文件
cd web && pnpm test -- --testPathPattern=page.test.tsx
cd server && pnpm test -- --testPathPattern=ai-service.test.js

RAG 系统测试

# 测试 RAG 准确性和质量评分
pnpm run rag:test-accuracy

# 运行综合 RAG 系统测试
pnpm run rag:test

# 验证加载数据的完整性和性能
pnpm run rag:verify

🤝 贡献

欢迎提交 Pull Request 和 Issue！

开发工作流程

Fork 项目
创建功能分支
提交更改
推送到分支
创建 Pull Request

📄 许可证