Design and development of financial applications using ontology-based multi-agent systems

Abstract

Researchers in the field of finance now use increasingly sophisticated mathematical models that require intelligent software on high performance computing systems. Agent models to date that are designed for financial markets have their knowledge specified through low level programming that require technical expertise in software, not normally available with finance professionals. Hence there is a need for system development methodologies where domain experts and researchers and can specify the behaviour of the agent applications without any knowledge of the underlying agent software. This paper proposes an approach to achieve the above objectives through the use of ontologies that drive the behaviours of agents. This approach contributes towards the building of semantically aware intelligent services, where ontologies are used rather than low level programming to dictate the characteristics of the agent applications. This approach is expected to allow more extensive usage of multi-agent systems in financial business applications.