A General Methodology for Internalising Multi-level Model Typing

Abstract

Multilevel Modelling approaches allow for an arbitrary number of abstraction levels in typing chains. In this paper, a transformation of a multi-level typing chain into a single all-covering representing model is proposed. This comprehensive model is of equal size as the most concrete model in the chain and encodes all typing information in its labels, such that the typing chain can completely be restored. This guideline for maintaining multi-level typing chains in respective implementations of multi-level typing environments is based on a categorical equivalence theorem, which we generalize to a more convenient graph-oriented version.