AI-Knowledge-Graph: Convert Unstructured Text into Interactive Knowledge Graphs 77

ai-Knowledge-Graph is a system that converts unstructured text documents into interactive knowledge graphs. It takes unstructured text, extracts information, and presents it in the form of Subject-Predicate-Object (SPO), ultimately displayed as an interactive graph.

AI-Knowledge-Graph Features

• Text Chunking: Splits large documents into smaller chunks for easier processing.

• Knowledge Extraction: Extracts SPO triples from text chunks to represent entities and relationships.

• Entity Standardization: Unifies different expressions of the same entity to avoid duplication.

• Relationship Inference: Discovers relationships between disconnected parts of the graph to make it more complete.

• Interactive Visualization: Generates interactive graphs that can be viewed in the browser.

• Multi-Service Support: Compatible with OpenAI API services (such as Ollama, LM Studio, etc.).

• Highly Configurable: Adjust models, API settings, etc. through configuration files.

• Command-Line Tool: Provides a command-line tool for convenient system execution.

• Modular Design: Easy to maintain and extend; users can modify or add modules.

• Statistics: Provides detailed graph statistics to help understand the structure.

AI-Knowledge-Graph Use Cases

• Academic Research: Quickly extract information from literature and build knowledge graphs.

• Enterprise Knowledge Management: Build internal knowledge bases to improve efficiency.

• Content Creation: Organize complex information and provide intuitive knowledge structures.

• Intelligent Search: Directly answer questions.

• Recommendation Systems: Perform precise recommendations.

• Q&A Systems: Map questions to graph queries.

• Financial Risk Control: Identify risk associations.

• Medical Diagnosis: Assist doctors in diagnosis and treatment recommendations.

How to Use AI-Knowledge-Graph

• Install Dependencies: Requires Python 3.11+. Install packages via pip install -r requirements.txt or uv sync.

• Configure Settings: Configure parameters in the config.toml file.

• Run the System: Use the following command to process text and generate the graph:

  python generate-graph.py --input your_text_file.txt --output knowledge_graph.html

• Or use UV:

  uv run generate-graph.py --input your_text_file.txt --output knowledge_graph.html

• View Results: Open the generated HTML file in a browser.

Project Link:https://github.com/robert-mcdermott/ai-knowledge-graph