MASEO: A Multi-Agent System for Explainable Ontology Generation

This repository provides the artifact for MASEO, a research-oriented multi-agent system that automated generate ontologies from competency questions, with a built-in focus on explainability. It aims to make the process of ontology generation more transparent, modular, and intelligent by distributing tasks among specialized agents. Each specialized agent is designed to keep track the logic behind each entity in generated ontology.

MASOE Structural Overview

The pipeline consists of four sequential stages:

Agent	Model	Responsibility
Ontology Generation Agent	deepseek-reasoner	Generates the initial OWL ontology from CQs
Syntax Repair Agent	deepseek-reasoner	Fixes RDF/XML syntax errors reported by the parser
Logical Consistency Agent	deepseek-reasoner	Repairs logical inconsistencies reported by HermiT
Pitfall Resolution Agent	deepseek-reasoner	Resolves ontology modeling pitfalls reported by OOPS!

The illustration of the MASOE framework:

Features

• End-to-end automation — from a list of CQs to a validated ontology
• Role-based agents — each stage is handled by a dedicated LLM agent with a specific instruction and responsibility
• Provenance tracking — every ontology entity carries an append-only vaem:rationale log attributed to the agent that made each change, and a dc:source log linking each change back to the CQ, pitfall, or error that motivated it

---

Installation

Python dependencies

pip install agno rdflib requests beautifulsoup4 pydantic

External tools requirements

Tool	Purpose	Setup
HermiT Reasoner	Logical consistency checking	Download HermiT.jar and update the path in reason_ontology()
OOPS! REST API	Ontology pitfall detection	No local setup required — uses the public REST endpoint
Java (JRE 8+)	Required to run HermiT	sudo apt install default-jre

Configuration

Make the changes in the agent_framework.py to ensure the correctness of variables.

# Path to your OOPS! request template
REQUEST_TEMPLATE = "src/templates/oops_request_template.xml"

# Path to HermiT JAR inside reason_ontology()
"java", "-jar", "/path/to/HermiT.jar"

Execution Command Line

Argument	Required	Description
--api_key	Yes	DeepSeek API key
--cqs_file	Yes	Path to a JSON file containing the list of competency questions
--save_file	Yes	Path where the final OWL ontology will be saved
--agent_method	No	true runs the full 4-stage pipeline; false runs generation only (default: true)

python cli.py \
    --api_key      YOUR_DEEPSEEK_API_KEY \
    --cqs_file     dataset/competency_questions/VGO.json \
    --save_file    dataset/generated_ontology/Gen_VGO.owl \
    --agent_method true

Input File Format

The input file that contains Competency Questions should be a JSON array of strings:

{
  "CQ1": "What is the genre of a game?",
  "CQ2": "Which players have purchased an in-app item?",
  "CQ3": "What is the username of a player?",
...
}

Output File Format

The internal representation and output format is available at our documentation [TBD]

Documentation

Additional documentation of the project is available at https://maseo.readthedocs.io/en/latest/?badge=latest

Acknowledgements

This work was supported by the grant SOEL: Supporting Ontology Engineering with Large Language Models PID2023-152703NA-I00 funded by MCIN/AEI/10.13039/501100011033 and by ERDF/UE. The authors would also like to thank the EDINT (Espacios de Datos para las Infraestructuras Urbanas Inteligentes) ontology development team for sharing the project resources for evaluation purposes.