🌿autoSKG

Automated Scientific Knowledge Graph Construction Pipeline for PlantScience.ai

autoSKG is the backend engine that powers PlantScience.ai, an LLM-powered virtual scientist for plant biology.
It automatically extracts, structures, and updates scientific knowledge from research papers to build a domain-specific, continuously evolving knowledge graph (KG).

---

🚀 Features

• Automated Knowledge Extraction – Converts unstructured scientific literature (PDFs) into structured entities and relationships.
• PlantScience.ai Integration – Forms the core KG engine supporting the PlantScience.ai virtual scientist.
• Continuous Updates – Supports incremental updates by detecting and integrating new knowledge from added PDFs.
• Traceable Knowledge – Every entity and relationship in the graph is linked to its original publication source.
• Dockerized Pipeline – Fully containerized for reproducible and scalable deployment.

---

📦Prerequisites

• Docker
• Docker Compose
• An OpenAI API key
• (Optional) A Docker Hub account if pushing/pulling images

💡 Place your PDF documents in the input_pdfs/ folder before running.

---

🐳Run with Docker Compose

1.Clone the repository

git clone https://github.com/COLA-Laboratory/autoSKG.git
cd autoSKG

2. Add input PDFs

Place your PDF files inside the input_pdfs/ directory:

input_pdfs/
├── paper1.pdf
└── paper2.pdf

3. Set your OpenAI API Key

The autoSKG requires an OpenAI API key for GPT-based processing.

1. Open the .env file (create one if it doesn't exist)
2. Add the following line:

OPENAI_API_KEY=your_openai_api_key_here

4. Start the pipeline

Run the following command to start all services:

docker compose up --build

---

📂 Output

After execution, results will be saved to:

kg_project/
├── input/ # Preprocessed text data extracted from PDFs
├── cache/ # Temporary data
└── output/ # Final structured knowledge graph (KG) and extracted entity-relationship files

---

🐙 Knowledge Graph Update

1. Add PDF files inside the input_pdfs/ directory
2. Start the pipeline Run the following command to start all services:

# Add new files to input_pdfs/
docker compose up --build

---

🧹 Clean Up

To stop and remove containers:

docker compose down

---

🌱 About PlantScience.ai

PlantScience.ai is a virtual scientist for plant biology, combining large language models with the domain-specific knowledge graph built by autoSKG. It provides accurate, citation-grounded answers and interactive graph visualizations of biological concepts — supporting next-generation scientific discovery.

License

This project is licensed under the Apache License 2.0.

© 2025 COLA Laboratory. All rights reserved.