“…In particular, PFs allow analysts to analyze complex problems by decomposing them into simpler ones; these basic problems are modeled according to basic patterns (i.e., the frames, which represent common, well understood problems); then it is necessary to (informally) show that the user requirements are satisfied by the outcome of the previously defined modeling activity; finally, the various problem frames are composed into a complete description. Meta-models for Problem Frames have been proposed by a few researchers [2], [3], [5], [6], [7], [10], including the authors of this paper [11]. However, all these meta-models (except, in a partial way [12]) do not address the (de)composition of problems, which is a fundamental aspect of the PF technique.…”