|
[[File:The OPM to Activity Diagram Map.jpg|centre|thumb|598x598px]]
==Study of model comprehension==
In order to assess and examine the effectiveness of the newly developed OPM-to-SysML algorithm and application, Dov Dori and Yariv Grobshtein performed a controlled experiment, which had two major goals:
# Since improved system comprehension was among the potential predicted benefits, they ventured to examine whether the additional SysML diagrams that had been generated automatically with their new application affected system model comprehension;
# They wished to test the quality of the autocreated diagrams, mainly in terms of modeling errors and inconsistencies with the original OPM model from which they were generated.
They carried out the experiment within a lecture of the course "Enterprise Systems Modeling" taught in 2008. Testing the null hypothesis that there is no difference in comprehension level between the OPM-only model and the combined OPM-and-SysML model, against the two-sided alternative that there is a difference, they rejected the null hypothesis in both cases. For both systems, the students' answers to comprehension questions relating the combined OPM-and-SysML models were significantly better than answers relating to OPM-only models.
It seems that the additional diagrams were helpful in answering some of the questions more than the others. Only one question of eight in each of the cases scored lower when the combined diagrams specification was used. The SysML diagram that seems to have the biggest added value is the Block Definition Diagram (BDD), which specifies system hierarchy and features of blocks like attributes and operations. The relatively high helpfulness of the BDD is evident from both the students'assessment of the contribution of this diagram type and from analysis of the scores in the individual questions in both systems. According to the results of the experiment, the Use Case diagram seems to have the least effect on system comprehension, and this is also in agreement with the students' assessment of the Use Case diagram type contribution. In between are the Activity diagram and the State Machine diagram, which appear to have a positive effect, albeit not as high as the effect of the BDD.
As noted, the other goal of the experiment was to assess the quality of the OPM-to-SysML application in terms of errors and inconsistencies between the source OPM model and the generated SysML diagrams. In this regard, the students did not find any issues of errors, inconsistencies, or contradictions associated with the SysML diagrams, indicating that the application has achieved a high level of model-to-model translation fidelity.
In the study an algorithm for automatic SysML views generation from an OPM model was developed and applied. For each SysML view, a mapping scheme from OPM elements to SysML elements constitutes the foundation for the OPM-toSysML sub algorithm and translation engine application. The mappings are partial and do not utilize all the available language elements. The software implementation was OPCAT, the OPM-supporting modeling tool.<ref>{{cite journal |last1=Grobshtein |first1=Yariv |last2=Dori |first2=Dov |title=Generating SysML views from an OPM model: Design and evaluation |journal=Systems Engineering |date=February 2011 |volume=14 |issue=3 |pages=327–340 |doi=10.1002/sys.20181 |ref=SysMLtoOPMEval }}</ref>
== See also ==
| 