Abstract-Multi-Agent System (MAS), and more specifically, ontology-based MAS, are increasingly being proposed and used within the medical domain. In this paper we represent an ontology-based multi-agent system specifically designed to intelligently retrieve information about human diseases. The human disease ontology is organized according to the four dimensions: disease types, symptoms, causes and treatments. The multi-agent system consists of four different types of agent: Interface, Manger, Information and Smart agent. We use of UML 2.1 to model social and goal-driven nature of agents. We believe that UML 2.1 has not only provided a way for standardized notation of MAS, but also for effective representation of the dynamic processes associated with these MAS.Keywords-UML, modeling of multi-agent systems, ontologybased multi-agent systems, information retrieval.