Skip to content

README.md

Latest commit

 

History

History
93 lines (69 loc) · 3.2 KB

README.md

File metadata and controls

93 lines (69 loc) · 3.2 KB

Chinese Poetry Web Application

This project is a Flask-based web application designed for managing, querying, and generating Chinese poetry. It integrates various features, including poetry data management, AI-powered poetry generation using OpenAI, and a knowledge graph powered by Neo4j to visualize relationships among poems, authors, and key themes extracted from the poems.

Features

  • Poetry Data Management:
    The application automatically loads poetry data from JSON files organized by category (e.g., Tang poetry, Song poetry). Users can search poems by title, author, or content.

  • Poetry Details & Statistics:
    View detailed information for individual poems along with statistics such as total poems, unique authors, and top authors based on the number of poems.

  • AI-Powered Poetry Generation:
    A dedicated chat page allows users to enter a prompt (theme or keyword) and generates a custom Chinese poem using OpenAI¡¯s GPT-based API.

  • Knowledge Graph Integration:
    The project builds a knowledge graph using Neo4j to capture relationships between:

    • Authors and the poems they created.
    • Poems and the keywords or themes they express.

    Users can update and query the knowledge graph through specific routes.

  • User-Friendly Navigation:
    A top navigation bar is available on all pages, allowing quick access to:

    • Home (Poetry Query)
    • AI-Powered Poetry Generation
    • Update Knowledge Graph
    • Query Knowledge Graph

Installation & Setup

Install Dependencies

Create a requirements.txt file containing:

  • Flask
  • py2neo
  • openai
  • jieba (optional)

Then run:

pip install -r requirements.txt

(Make sure to run this command in your virtual environment.)

Configure Environment Variables

Set your OpenAI API key to os env:

export OPENAI_API_KEY="your_openai_api_key"

Configure Neo4j

Install and run Neo4j from https://neo4j.com/download/.
Update the Neo4j connection details in the kg.py and app.py files (replace "your_password" with your actual password).

Run the Application

Run:

python app.py

Then, open your web browser and navigate to http://127.0.0.1:5000.

Usage

Poetry Query:

On the homepage, select a poetry category and optionally enter a search keyword. View statistics and click "View Details" to read a full poem.

Poem Detail:

Access detailed information about a poem, including its title, author, and full text.

AI Poetry Generation:

Navigate to the "AI Poetry Generation" page to enter a prompt and receive an AI-generated poem.

Knowledge Graph:

  • Update Knowledge Graph: Visit /update_kg to build or update the Neo4j knowledge graph.
  • Query Knowledge Graph: Use /query_kg (e.g., by author) to explore relationships in the graph.

Navigation

The application features a consistent top navigation bar on all pages, providing quick links to:

  • Home (Poetry Query)
  • AI-Powered Poetry Generation
  • Update Knowledge Graph
  • Query Knowledge Graph

License

This project is licensed under the MIT License.

Acknowledgements

  • Flask
  • Neo4j
  • OpenAI
  • Contributions from various open-source projects in the fields of Chinese poetry, natural language processing, and knowledge graphs.