| --- |
| title: Scripture Knowledge Graph |
| emoji: ✨ |
| colorFrom: blue |
| colorTo: purple |
| sdk: gradio |
| sdk_version: "6.20.0" |
| app_file: app.py |
| pinned: true |
| license: mit |
| short_description: Bible notes to an AI knowledge graph with swarm intelligence |
| --- |
| |
| # Scripture Knowledge Graph |
|
|
| Transform your Bible study notes into an interactive knowledge graph powered by graph theory, swarm intelligence, and AI theological analysis. |
|
|
| ## Features |
| - **NLP entity extraction** — automatically identifies people, places, themes, theological concepts, books, verses, time periods, and events, plus directed relationships ("Paul wrote to Timothy") |
| - **Interactive D3 force graph** — drag, zoom, click to explore; hexagon/diamond/star/triangle node shapes by entity type; neon glow rendering |
| - **Graph algorithms (NetworkX)** — PageRank, greedy-modularity communities, betweenness/degree/eigenvector centrality, clustering coefficient, diameter, density, bridges, articulation points |
| - **ACO Swarm Intelligence** — a real Ant Colony System (Dorigo & Gambardella, 1997) traces the dominant conceptual pathway through your notes, with a live convergence readout |
| - **AI Insights** — Claude streams a theological + graph-structural analysis of your notes, grounded in the computed topology and swarm results |
|
|
| ## How to use |
| 1. Paste your Bible study notes (or click **Sample**) |
| 2. Click **Build Knowledge Graph** |
| 3. Explore — click nodes, try layout modes, drag, zoom |
| 4. Click **Run Swarm Analysis** to let ants find the dominant pathway |
| 5. Open the **AI Insights** tab and click **Generate Insights** |
|
|
| ## Setup |
|
|
| This Space calls the Anthropic API server-side. Add your key as a **Space Secret** named `ANTHROPIC_API_KEY` (Settings → Variables and secrets). It is never exposed to the browser. |
|
|
| ## Technical stack |
| Graph algorithms: NetworkX · Visualization: D3.js v7 · AI: Claude (Anthropic, server-side streaming) · Swarm: Ant Colony System (Dorigo & Gambardella, 1997) |
|
|
