Data Mining MCP Server

A Model Context Protocol (MCP) server for managing and querying Data Mining course materials. This server provides intelligent access to your lecture notes, tutorials, tutorial answers, and past papers through AI-powered tools.

📚 Features

• Search Across All Materials: Full-text search across lectures, tutorials, answers, and past papers with PDF/DOCX support
• Week-Based Organization: Retrieve all materials for any specific week (1-13)
• Past Paper Access: Quick access to exam papers from 2015-2024
• Course Overview: Get a complete overview of available materials
• PDF & DOCX Support: Instant text extraction using PyMuPDF and python-docx
• Resource Access: Direct access to any document by path

🚀 Quick Start

Prerequisites

• Node.js 18 or higher
• Python 3.10 or higher
• npm

Installation

1. Clone or navigate to this directory
2. Install Node.js dependencies:

npm install

3. Set up Python environment:

python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
pip install PyMuPDF python-docx

4. Build the project:

npm run build

Adding Your Course Materials

Place your materials in the course-materials/ directory following this structure:

course-materials/
├── lectures/week-1/          # Week 1 lecture notes
├── lectures/week-2/          # Week 2 lecture notes
├── ...
├── lectures/week-13/         # Week 13 lecture notes
├── tutorials/week-1/         # Week 1 tutorial questions
├── tutorial-answers/week-1/  # Week 1 tutorial answers
├── additional-resources/     # Supplementary materials
└── past-papers/              # Past year papers by year (2015-2024)
    ├── 2015/
    ├── 2016/
    ├── ...
    ├── 2023/
    └── 2024/

Supported file formats:

• PDF (.pdf) - with instant text extraction
• Word Documents (.docx)
• Markdown (.md)
• Text (.txt)
• Any text-based format

Example: Adding Week 1 Materials

# Add lecture notes
cp ~/Downloads/week1-classification.pdf course-materials/lectures/week-1/

# Add tutorial
cp ~/Downloads/week1-tutorial.pdf course-materials/tutorials/week-1/

# Add tutorial answers
cp ~/Downloads/week1-answers.docx course-materials/tutorial-answers/week-1/

# Add past papers by year
cp ~/Downloads/2023-midterm.pdf course-materials/past-papers/2023/
cp ~/Downloads/2023-final.pdf course-materials/past-papers/2023/
cp ~/Downloads/2024-exam.pdf course-materials/past-papers/2024/

🔧 Using the Server

With VS Code GitHub Copilot (Recommended)

1. Open VS Code Command Palette (Cmd+Shift+P on macOS, Ctrl+Shift+P on Windows/Linux)
2. Type and select: "MCP: Open User Configuration"
3. Add this configuration to your mcp.json:

{
  "servers": {
    "data-mining": {
      "type": "stdio",
      "command": "node",
      "args": ["/ABSOLUTE/PATH/TO/Data Mining MCP Server/build/index.js"]
    }
  }
}

Replace /ABSOLUTE/PATH/TO/ with the actual path to this directory.

4. Save the file and restart VS Code or GitHub Copilot

Note: The MCP configuration file is located at:

• macOS/Linux: ~/Library/Application Support/Code/User/mcp.json
• Windows: %USERPROFILE%\AppData\Roaming\Code\User\mcp.json

With Claude Desktop (Alternative)

If you prefer Claude Desktop, add to ~/Library/Application Support/Claude/claude_desktop_config.json on macOS:

{
  "mcpServers": {
    "data-mining": {
      "command": "node",
      "args": ["/ABSOLUTE/PATH/TO/Data Mining MCP Server/build/index.js"]
    }
  }
}

With MCP Inspector (for testing)

npx @modelcontextprotocol/inspector node build/index.js

🛠️ Available Tools

1. search-materials

Search across all course materials by content or filename.

Parameters:

• query (string): Search term

Example:

Search for "decision trees" in all materials
Search for "clustering" in materials

2. get-week-materials

Retrieve all materials for a specific week.

Parameters:

• week (string): Week identifier (e.g., "week-1", "week-2", ..., "week-13")

Example:

Get lecture notes for week 5
Get tutorial answers for week 3

3. list-past-papers

List all available past year exam papers organized by year (2015-2024).

Parameters: None

Example:

Show me all past papers
List available exam years

4. get-year-papers

Get all past papers for a specific year.

Parameters:

• year (string): Year to retrieve papers from (e.g., "2015", "2024")

Example:

Get all papers from 2024
Show me 2023 exam papers

5. get-course-overview

Get an overview of all available course materials.

Parameters: None

Example:

Give me an overview of the course
What materials are available?

6. read-file

Read and extract text from PDF or DOCX files.

Parameters:

• filePath (string): Relative path from course-materials folder
• maxPages (optional number): Limit pages to extract from PDFs

Example:

Read the week 3 lecture PDF
Extract text from lectures/week-5/decision-trees.pdf

📖 Resources

The server also exposes documents as resources that can be directly referenced:

• datamining://materials/week-{X} - Access materials for week X (1-13)
• datamining://past-papers/{YEAR} - Access past papers for specific year

�� Example Workflows

Studying for Week 3

1. Ask your AI assistant: "Help me study Week 3"
2. The AI will use get-week-materials to retrieve lectures, tutorials, and answers
3. Ask follow-up questions about specific concepts like "decision trees"

Searching for a Topic

1. Ask: "Find all materials about classification algorithms"
2. The AI will use search-materials to find relevant content across all weeks
3. Get direct access to the relevant documents

Exam Preparation

1. Ask: "Help me prepare for my Data Mining exam"
2. The AI will:
   • List all past papers using list-past-papers
   • Search for key concepts
   • Help you practice questions
   • Create a study plan based on available materials

🏗️ Development

Project Structure

.
├── src/
│   └── index.ts          # Main MCP server implementation
├── scripts/
│   └── extract_text.py   # PDF/DOCX text extraction utility
├── course-materials/     # Your course materials
│   ├── lectures/
│   ├── tutorials/
│   ├── tutorial-answers/
│   ├── additional-resources/
│   └── past-papers/
├── build/                # Compiled JavaScript (generated)
├── package.json
├── tsconfig.json
└── README.md

Making Changes

1. Edit src/index.ts
2. Rebuild: npm run build
3. Restart your MCP client to pick up changes

Watch Mode (for development)

npm run watch

This will automatically rebuild when you make changes to the source code.

�� Tips for Exam Preparation

1. Organize Your Materials: Keep materials organized by week for easy retrieval
2. Use Descriptive Filenames: Name files clearly (e.g., week3-decision-trees.pdf)
3. Combine with AI: Ask the AI to quiz you, explain concepts, or create study guides
4. Search Effectively: Use specific terms to find relevant materials across all weeks
5. Practice with Past Papers: Use the list-past-papers tool to access previous exams

🔒 Privacy

All materials are stored locally on your machine. The MCP server only provides access to your AI assistant through the Model Context Protocol. Your course materials never leave your computer.

📝 License

MIT

🆘 Troubleshooting

Server won't start

• Make sure you've run npm install and npm run build
• Check that Node.js version is 18 or higher: node --version
• Verify Python dependencies are installed in .venv/

Materials not showing up

• Verify files are in the correct directory structure under course-materials/
• Check file permissions (files must be readable)
• Try using the get-course-overview tool to see what's detected

Search not finding content

• Ensure PDFs have text layers (not scanned images without OCR)
• Check file encoding (should be UTF-8)
• Try searching for exact words from the document

PDF/DOCX extraction failing

• Verify Python virtual environment is activated
• Ensure PyMuPDF and python-docx are installed: pip list | grep -E "PyMuPDF|python-docx"
• Check Python path in your environment

📚 Example Materials Template

Create well-organized materials with clear naming:

course-materials/
├── lectures/week-1/
│   ├── w1-introduction-to-data-mining.pdf
│   └── w1-lecture-notes.md
├── tutorials/week-1/
│   └── w1-tutorial-questions.pdf
└── tutorial-answers/week-1/
    └── w1-tutorial-solutions.docx

🤝 Contributing

This is a study tool for Data Mining students. Feel free to customize it for your own needs or contribute improvements!

📖 Resources

• MCP Documentation
• MCP TypeScript SDK
• PyMuPDF Documentation

---

Happy Studying! 🎓 Use this tool to ace your Data Mining exam!