Designing an ontology-based multi-agent system for supply chain performance measurement using graph traversal

Abstract

Previous studies have not provided any information system (IS) solution for the supply chain (SC) performance measurement based on the Global Supply Chain Forum (GSCF) reference model due to the complexities of interoperability among the ISs in the SC. As an innovation, this paper has proposed an architecture that allows the SC Performance Measurement System (SCPMS) components to have multi-level interoperability by integrating the multi-agent technology as well as ontology technology. It has also proposed a method to develop the required ontologies for the PMS agents and developed the ontologies in the Web Ontology Language (OWL) to provide the semantic level interoperability. The syntax-level interoperability has been provided by a number of agent interaction protocols (AIPs). As another innovation, the authors have developed the SCPMS algorithms for the link analysis method by the Breadth-First Search (BFS) method and proved those algorithms. The authors have created the PMS agents based on those algorithms and prototyped a Multi-Agent System (MAS) for the SC of a tile manufacturer as a case study. The proposed multi-level architecture supports the GSCF model in the SC performance measurement, and improves the interoperability and the integration of the SCPMS.