VITAMIN: A Compositional Framework for Model Checking of Multi-Agent Systems

Research output: Contribution to journalConference articlepeer-review

Abstract

The verification of Multi-Agent Systems (MAS) poses a significant challenge. Various approaches and methodologies exist to address this challenge; however, tools that support them are not always readily avail able. Even when such tools are accessible, they tend to be hard-coded, lacking in compositionality, and challenging to use due to a steep learning curve. In this paper, we introduce a methodology designed for the formal verification of MAS in a modular and versatile manner, along with an initial prototype, that we named VITAMIN. Unlike existing verification methodologies and frameworks for MAS, VITAMIN is constructed for easy extension to accommodate various logics (for specifying the properties to verify) and models (for deter mining on what to verify such properties).

Original languageEnglish
Pages (from-to)648-655
Number of pages8
JournalInternational Conference on Agents and Artificial Intelligence
Volume1
DOIs
Publication statusPublished - 1 Jan 2025
Event17th International Conference on Agents and Artificial Intelligence, ICAART 2025 - Porto, Portugal
Duration: 23 Feb 202525 Feb 2025

Keywords

  • Formal Verification
  • Model Checking
  • Multi-Agent Systems
  • Software Engineering

Fingerprint

Dive into the research topics of 'VITAMIN: A Compositional Framework for Model Checking of Multi-Agent Systems'. Together they form a unique fingerprint.

Cite this