Unified modeling language (UML) is commonly employed as a general-purpose notation for software modeling. A number of extensions have been proposed to model a particular domain-specific application. UML Web application extension (WAE) is an extension dedicated to modeling Web applications.
Assessing the practicality of such an extension is not trivial. The classic way to measure the value of a graphical notation is to conduct an empirical study over a number of distinct user groups. The authors of this study found that “it is not possible to observe any significant benefit” associated with using WAE. However, “the gap between subjects with low skill or experience and highly skilled or experienced subjects” is significantly reduced.
After having read the paper, I still did not understand the reason for this lack of apparent benefit when using WAE. I also would have liked to see which other kinds of UML extensions the conclusions are applicable to.
The authors are making their claim based on an empirical study that involves solely academics and students. Therefore, their result may not be representative in an industrial setting. This is a common critique of such studies, as the authors themselves recognize.
Overall, Ricca et al. deliver a rigorous, objective, and critical observation of a stereotype-based UML extension.