Skip to content

hazelian0619/wordcloud2tester

Repository files navigation

WordCloud Emergence

Between two concepts, find the third that doesn't yet exist.

GAT-Powered Semantic Network Exploration - From fuzzy intuition to precise discovery


🌟 What It Is

WordCloud Emergence is a creative tool and a way of thinking.

It won't tell you "dragon + girl = what," but unfolds a fuzzy territory where concepts converge: Medusa, Xiaolongnü, Maleficent, Jade Fox...

At the intersections of these neighbors, you'll find the answer that belongs only to you.


💡 Why It Exists

Two Scenes

Scene One: The Classroom

A student asks: "What is justice?"

A good teacher doesn't answer directly. She says: "There's a story about a watchman..."

The student begins to explore: guard duty, responsibility, solitude, promise... Among these fuzzy neighbors, they walk toward their own understanding of justice.

Scene Two: The Creative Studio

Versace, Year of the Dragon campaign.

The creative director won't simply design "a dragon with a girl."

They write on the whiteboard: Dragon · Girl

Then ask: "In all of human cultural memory, who lives in both words?"

The team begins listing fuzzy neighbors:

  • Medusa — Snake-haired gorgon, echoing dragon's reptilian form
  • Xiaolongnü — Cool serenity and mysterious power in Eastern wuxia
  • Maleficent — Dark queen transforming into dragon in Sleeping Beauty
  • Jade Fox — Rebellious soul and weapon in Crouching Tiger, Hidden Dragon
  • The Girl with the Dragon Tattoo — Violent aesthetics and revenge in Nordic noir
  • Chihiro — Girl growing before the white dragon god in Spirited Away
  • Nezha — Gender-fluid dragon prince embodying rebellion and rebirth

The room goes quiet.

Creativity doesn't lie in the direct stacking of "dragon + girl," but at the intersections of these neighbors.


🧠 Core Philosophy

Fuzziness Guides Precision - We're not doing simple keyword matching. Through AI's "fuzzy understanding" capability, we guide users to discover precise connections between concepts. Just like human thinking: starting from fuzzy intuition, gradually crystallizing into concrete conceptual networks.

Fuzzy Neighbors

The best answers don't lie within concepts themselves.

They exist in the spaces between, where fuzzy neighbors dwell.

In teaching, this is Socratic questioning—stories that spark deeper thought.

In creativity, this is concept topology—surprises found at intersections.

Emergence

"Dragon + Girl" stacked together is dull.

But create distance, and something else appears between.

Medusa brings serpentine danger. Xiaolongnü brings Eastern serenity. Maleficent brings the possibility of transformation. Jade Fox brings rebellious tension.

Students reach their own understanding here. Creatives find unexpected connections here. Writers see images that don't yet exist here.

This is emergence.


🎯 How It Works

Creative Process

Step 1: Extract Core Concepts

Dragon + Girl

Step 2: Expand Neighbors in Knowledge Graph

What connects both dragon and feminine?

  • Medusa — Gorgon with serpentine hair, echoing reptilian forms
  • Xiaolongnü — Cool serenity and mysterious power in Eastern wuxia
  • Maleficent — Dark queen transforming into dragon in Sleeping Beauty
  • Jade Fox — Rebellious soul and weapon in Crouching Tiger, Hidden Dragon
  • The Girl with the Dragon Tattoo — Violent aesthetics and revenge in Nordic noir
  • Chihiro — Girl growing before the white dragon god
  • Nezha — Gender-fluid dragon prince, rebellion and rebirth

Step 3: Find Unique Creativity at Intersections

These fuzzy neighbors are where true inspiration lives.


⚙️ Technical Implementation

Phase 1: Fuzzy Semantic Understanding

  • AI receives user's fuzzy concepts
  • Deep learning models understand semantic boundaries
  • Generate multiple possible semantic directions

Phase 2: Attention Weight Calculation

  • Calculate attention weights for each related concept
  • Weight = Semantic Similarity × Concept Importance
  • Form weighted conceptual networks

Phase 3: Precise Path Discovery

  • User clicks trigger new semantic expansion
  • System records complete semantic paths
  • From fuzzy starting point to precise destination

🛠️ Technology Stack

Backend (Python)

  • Framework: FastAPI (inspired by Creative-Writing migration)
  • AI/ML: OpenAI API, PyTorch (GAT Networks)
  • Async: asyncio for concurrent processing
  • APIs: RESTful endpoints for concept expansion

Frontend (JavaScript)

  • Visualization: D3.js for force-directed graphs and word clouds
  • UI: Vanilla JS with modern CSS
  • Internationalization: i18n support for multiple languages
  • Responsive: Mobile-friendly design

Infrastructure

  • Deployment: Vercel (serverless)
  • Containerization: Docker support (from Creative-Writing)
  • Monitoring: Basic logging and error handling
  • Data Storage: JSON-based concept databases

Key Components Migrated from Creative-Writing

  • API Architecture: FastAPI patterns for scalable backends
  • Visualization Techniques: Advanced D3.js implementations for cognitive state displays (adapted for word cloud weights)
  • Deployment Configurations: Docker and Kubernetes setups for production
  • Knowledge Graph Concepts: Graph-based semantic expansion algorithms

✨ Core Features

  • 🧠 Fuzzy Semantic Understanding - AI's "fuzzy understanding," not exact matching
  • Attention Weights - Concept importance visualized through weights
  • 🔄 Path Tracking - Record complete journey from fuzzy to precise
  • 📊 Dynamic Layout - Weight-based force-directed graph layout
  • 🎯 Spotlight Mode - Focus on current exploration path

🚀 Try It Online

Experience Now →


🎯 Technical Demonstrations

Space-Time Exploration

Input: Space-Time

Fuzzy Understanding: time, space, dimension, relativity, travel, parallel universe, quantum, eternity

Attention Weight Analysis:

  • Relativity (0.95) - Core physics theory
  • Dimension (0.90) - Spatial concept
  • Travel (0.85) - Sci-fi imagination
  • Parallel Universe (0.82) - Theoretical physics
  • Quantum (0.80) - Microscopic world

Precise Path Discovery:

Space-Time → Relativity → Einstein → E=mc²
Space-Time → Dimension → 4D Space → Spacetime Curvature

📁 Project Structure

wordcloud2tester/
├── backend/           # Python FastAPI backend
│   ├── path-expand.py # Concept expansion logic
│   ├── test.py       # API tests
│   └── requirements.txt
├── frontend/          # Web frontend
│   ├── index.html    # Main page
│   ├── src/          # JavaScript sources
│   └── i18n-translations.js
├── config/            # Configuration files
│   └── vercel.json   # Vercel deployment config
├── data/             # Data files and assets
├── docs/             # Documentation
│   ├── README.md     # Detailed docs
│   └── DEPLOY_GUIDE.md
└── README.md         # This file

🏃‍♂️ Quick Start

Prerequisites

  • Python 3.8+
  • Node.js 16+ (for development)
  • OpenAI API key

Installation

  1. Clone the repository

    git clone https://github.com/hazelian0619/wordcloud2tester.git
    cd wordcloud2tester
  2. Setup backend

    cd backend
    pip install -r requirements.txt
    # Set your OpenAI API key
    export OPENAI_API_KEY="your-key-here"
  3. Run locally

    # Start backend
    python path-expand.py
    
    # Open frontend in browser
    cd ../frontend
    python -m http.server 8000
    # Visit http://localhost:8000

Deployment

  • Vercel: Automatic deployment from main branch
  • Docker: See deployment guide in docs/DEPLOY_GUIDE.md

🤝 Contributing

This project incorporates elements from the Creative-Writing repository (https://github.com/hazelian0619/Creative-Writing) to enhance functionality, particularly in API architecture, visualization techniques, and deployment practices.

Contributions are welcome! Please feel free to submit a Pull Request.


📄 License

MIT License - see LICENSE file for details

About

模糊度引导精确性:AI语义词云

Resources

License

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors