Publication Date

Winter 12-7-2014


Proceedings of the 2014 Winter Simulation Conference


While domain specific languages are well established for describing the system of interest in modeling and simulation, the last years have seen increasingly domain specific languages also exploited for specifying experiments. This development, whose application areas range from computational biology to network simulation, is motivated by the desire to facilitate the reproducibility of simulation results. Thereby, the experimentation process is treated as a first class object of simulation studies. As the experimentation process contains different tasks such as configuration, observation, analysis, and evaluation, domain-specific languages can be exploited to specify experiments as well as individual sub-tasks or even the goal of the experiment, thus opening up new avenues of research. The focus of our discussion will be on what information to express, also based on existing approaches. Referring to how to express the required information, we will sketch some of the pros and cons of external and embedded domain specific languages.