The JAS-mine platform‎ > ‎Concept‎ > ‎

Input-Output communication

Data management is a major factor to be weighed in for the creation of a simulation tool. Building on the vast number of software solutions available, JAS-mine allows the user to separate data representation and management from the implementation of processes and behavioral algorithms.

The main value added of the platform lies in the integration with relational database management systems (RDBMS) through ad-hoc Java libraries. The management of input data persistence layers and simulation results in JAS-mine is performed using standard database management tools, and the platform takes care of the automatic translation of the relational model of the database into the object-oriented simulation framework thanks to an ORM layer.

This also allows to separate data creation from data analysis, which is crucial for understanding the behavior of the simulation model. As the statistical analysis of the model output is possibly intensive in computing time, performing it in real time might be an issue, in large-scale applications. A common solution is to limit it on a selected subset of output variables. This however requires identifying the output of interest before the simulation is run. If additional computations are required to better understand how the model behaves, the model has to be run again: the bigger the model, the more impractical this solution is.

On the other hand, the power of modern RDBMS make it feasible to keep track of a much larger set of variables, for later analysis. Also, the statistical techniques envisaged, and the specific modeler’s skills, might suggest the use of external software solutions, without the need to integrate them in the simulation machine.

Finally, keeping data analysis conceptually distinct from data production further enhances the brevity, transparency and clarity of the code.