| title: Concept Knowledge Graph Builder | |
| emoji: 🕸️ | |
| colorFrom: indigo | |
| colorTo: purple | |
| sdk: gradio | |
| python_version: "3.10" | |
| app_file: app.py | |
| pinned: false | |
| license: mit | |
| # Concept Knowledge Graph Builder | |
| This is a premium, lightweight, interactive Concept and Entity network mapping tool designed specifically for computational social science and humanities courses. | |
| ## Features | |
| - **Flexible Data Input**: Paste raw text directly or upload datasets (`.csv`, `.xlsx`, or `.txt`). | |
| - **Flexible Network Extraction Modes**: | |
| - **Local Noun-Chunk Parser (CPU & Fast)**: Runs high-speed extraction utilizing SpaCy dependency mapping and noun chunks to extract key ideas/entities and calculate sentence-level co-occurrences locally on CPU. | |
| - **Transformers (AI Mode)**: Leverages advanced generative models (like `Qwen2.5`) via Hugging Face's Serverless API with your personal access token to generate structured relational triple JSON structures. | |
| - **Beautiful Graph Visualizations**: Renders complete, fully responsive interactive 2D circular network graphs using Plotly Scatter Network layers natively. | |
| - **CSV Data Export**: Easily export extracted node tables and edge relationship files as standard CSV files. | |