Open-Source Knowledge Graph Builder Supports Multi-Backend LLMs and Visualization
tom_doerr · x · 2026-08-09
An open-source modular system named Knowledge Graph Builder (KGB) has been released on GitHub to extract knowledge graphs from text. Key features of the project include:
- Multi-Backend Support: Compatible with Gemini API and local models like Ollama and LM Studio.
- Structured Extraction: Extracts source-grounded triples (head-relation-tail) via langextract.
- Graph Augmentation: Offers iterative strategies to bridge disconnected graph components, tracking whether triples are explicitly extracted or contextually augmented.
- Interactive Visualization: Integrates Cytoscape.js to generate interactive network graphs with node dragging, search/filter, and entity highlighting.
More from coding & agent
- LLMs Master AppleScript: Agents Can Now Drive Native Apple Apps — xeophon · 2026-08-09
- Developer Insight: AI Software Isn't One-and-Done, Agent Upkeep Costs Are Significant — shakoistsLog · 2026-08-09
- Dev Ports Command & Conquer: Red Alert 2 to iPhone Using AI — ammaar · 2026-08-09
- Standardizing the Agent Stack: Independent Memory Components to Become the Norm — hwchase17 · 2026-08-09
- MCP tool calls pose context bloat risks; developer proposes pipe-filter solution — gnoraz_theorc · 2026-08-09
- Dev Tool: env-run Simplifies Environment Variable Management — Rasmic · 2026-08-09