Semantic multi-agent system to assist business integration: An application on supplier selection for shipbuilding yards

Abstract To compete in the fierce market, a key issue for shipbuilding companies is to effectively integrate outer resources with the inner business. However, multi-agent systems (MAS) developed so far have proven to be inapt to support the whole process of making a coordinated supply plan. Meanwhile, huge amount of data emerging from various kinds of enterprise infrastructures have not been semantically consolidated for smooth and efficient workflows. To overcome these barriers, this paper proposes a hybrid semantic-augmented MAS (OMAS), consisting of a Service Management Agent, a Configuration Agent, a Member Management Agent, a Process-Driven Module and a Semantic-Enhanced Module. These components collaborate to integrate heterogeneous data devoid of semantic ambiguities, thereby improving workflows’ coherence and efficiency. A scalable ontology grid is initialized as the operational foundation after analyzing the related works. In addition, OMAS innovatively adds semantic assistance into the supplier selection process wherein a two-phase negotiation algorithm is suggested. Finally, an illustrative example from a marine company is conducted to demonstrate the benefits of OMAS, through which the effectiveness of the proposed algorithm is proved to be much better than that of a contemporary contribution.