Semantic mashups for simulation as a service with tag mining and ontology learning

Nowadays, there is a trend for delivering the Simulation as a Service using web-based/cloud-based services. Existing simulation services cannot be easily discovered and composed. Although semantic mashups have become popular for implementing service composition in the Web 2.0, there are yet no semantic mashups applications focusing on modeling and simulation. Here, we propose the first existing layered architecture based on semantic mashups improving the composition of Simulation as a Service. Besides, we propose using ontology learning and tagging systems to avoid pre-defined ontology efforts and to increase the automation of composition through user participation. The general idea is to mine tag signatures from the userinterested simulation-related services automatically, to generate a tag ontology tree from the mined tag signatures automatically, and then to compose the services based on the learnt tag tree ontology. This unique approach for simulation services mashups can boost the reusability, integration, interoperability of Simulation as a Service.