kirancodes.me
To Proof Maintenance & Beyond!

Enabling dynamic metamodels through constraint-driven modeling

Andreas Demuth

Abstract

Metamodels are commonly used in Model-Driven Engineering to define available model elements and structures. However, metamodels are likely to change during development for various reasons like requirement changes or evolving domain knowledge. Updating a metamodel typically leads to non-conformance issues with existing models. Hence, evolution strategies must be developed. Additionally, the tool implementation must also be updated to support the evolved metamodel. We propose the use of metamodel-independent tools with unified modeling concepts for working with all kinds of metamodels and models. By applying the Constraint-Driven Modeling approach and generating model constraints from metamodels automatically, we solve the described issues and enable dynamic, evolving metamodels. A prototype implementation has shown the feasibility of the approach and performance tests suggest that it also scales with increasing model sizes.

BibTeX
@inproceedings{Demuth:ICSE12,
  author    = {Andreas Demuth},
  title     = {Enabling dynamic metamodels through constraint-driven modeling},
  booktitle = {ICSE},
  pages     = {1622--1624},
  publisher = {{IEEE} Computer Society},
  year      = {2012},
}

Related papers