Semantic selection for model composition using SAMSaaS

The reuse of existing models in modeling and simulation software is becoming increasingly important. One way to improve model reuse is through composition. Even though much work has been done for model composition at the syntactic level, research at the semantic level is still under development. In particular, selection for model composition based on semantics is complex, because it is hard for casual users to understand the formal definition of models. Likewise, there are no automated mechanisms to build domain-specific ontologies based on the models. In order to deal with the above issues, we introduced SAMSaaS (Semantic Architecture for Modeling and Simulation as a Service), a layered architecture to deploy and compose M&S resources as services. Here, we focus on the upper layers of the architecture, and present a tag mining method and a tag-tree ontology learning approach to learn domain specific ontology for the models. We show an application to DEVS examples to show its success and possible applications.