A formalisation of deep metamodelling

Abstract

Esta es la versión de autor de la comunicación de congreso publicada en: This is an author produced version of a paper published in: Abstract. Metamodelling is one of the pillars of model-driven engineering, used for language engineering and domain modelling. Even though metamodelling is traditionally based on a two-metalevel approach, several researchers have pointed out limitations of this solution and proposed an alternative deep (also called multi-level) approach to obtain simpler system specifications. However, this approach currently lacks a formalisation that can be used to explain fundamental concepts such as deep characterisation, double linguistic/ontological typing and linguistic extension. This paper provides such a formalisation based on the Diagram Predicate Framework, and discusses its practical realisation in the METADEPTH tool.

DeepTelos: Multi-level Modeling with Most General Instances

“…Adding to a metaclass c an MGI m results in a base class m with a power type c. The contribution of DeepTelos is to fully integrate this construct in a metamodeling language and system which already comes with support for unbounded deep characterization via 'mediated' [20] properties and full support for metaclasses. This also sets DeepTelos apart from work on the powertype pattern [4] where the powertype role is played by a 'normal' class.…”

“…We take a perspective that is inverse to power types, because our starting point are metaclasses in Telos. With MGIs we add full support for deep characterization to Telos, making it possible that the metaclass introduces (via its MGI) a property that is directly instantiated by the individual (its instance-instance, i.e., potency 2) without an intermediate instantiation step at the class -this is similar to what Rossini [20] refers to as 'semantics of single-potency'.…”

“…Early approaches were materialization [19] and power types [18]. Many current approaches [6,20,12,16] to multi-level modeling assign potencies to relationships, attributes, and classes. Potencies constrain how many times the respective concept can be instantiated until it can no further be instantiated.…”

Co-evolving meta-models and their instance models: A formal approach based on graph transformation

“…Graphs and graph-like structures have been successfully used for representing software models and meta-models [4,[22][23][24]. Basic meta-modeling concepts can be represented by typed graphs.…”

A Comparison of Two-Level and Multi-level Modelling for Cloud-Based Applications

“…A potency is a natural number (or zero) indicating at how many levels an element can be instantiated (cf. [21] for a formal discussion of different types of potency). In Fig.…”