immartian / wakangogen

Using LLMs to Research Chinese Language Terms Originated from Japanese

Wakan Gogen ("和漢語源")

Overview

This project is an interactive taxonomy bubble visualization built using FastAPI and D3.js. It includes a backend server for managing taxonomy data stored in MongoDB and a frontend for visualizing the data.

Features

• Interactive taxonomy bubbles using D3.js
• FastAPI backend for managing data
• MongoDB for data storage
• CORS enabled for API access

Requirements

• Python 3.8+
• MongoDB
• Docker (optional, for containerized deployment)

Running

1. Run with Python

uvicorn main:app --reload

Navigate to http://127.0.0.1:8000 to see the application running.

2. Run the Docker Container

sudo docker build -t wakangogen .
sudo docker run --env-file .env -p 8000:8000 wakangogen

---

Todos

• Collect a comprehensive list of Chinese terms with Japanese origins -> 452 collected terms
• Implement LLMs for taxonomy identification -> 7 taxonomies confirmed
• Develop a system for atomical analysis of terms -> WIP
• Create a relational database for storing term data -> done
• Design interactive visualizations for term relationships -> MVP
• Develop a web platform for exploring and contributing to the data