Metamodels are at the core of any model-driven projects and give the experts a linguistic means to abstractly represent the problem domain’s instances. Metamodels are prone to modifications due to new insights emerging from the domain, improvements, and features being added to the modeling language over time. Metamodel evolution may induce severe repercussions over the related artifacts, which might need to be consistently adapted. The co-evolution problem poses another issue related to the epistemic uncertainty arising as a response to a metamodel change where many different consistency restoration procedures are possible, e.g., restoring models and transformations. In this work we rely on the notion of information loss (IL) to understand which migration combinations are preferable to others in coupled evolution, by offering a ranking of the possible solutions. The IL denotes that part of the knowledge contained in the source models cannot be conveyed or translated into the target models, using a selected transformation, resulting in a loss of information. Such aspect represents an extra-functional quality that can be used for ranking the selection of certain transformation migrations in the context of multiple available alternatives and in combination with model migrations. Information loss induced by a (migrated) transformation can be defined as the amount of information lost executing the (migrated) transformation compared to the original output model obtained by executing the original transformation. The proposed approach is supported by a prototype tool that we demonstrate and validate.

Uncertainty management with extra-functional qualities in multi-artefact co-evolution

Basciani Francesco;Iovino Ludovico;
2021-01-01

Abstract

Metamodels are at the core of any model-driven projects and give the experts a linguistic means to abstractly represent the problem domain’s instances. Metamodels are prone to modifications due to new insights emerging from the domain, improvements, and features being added to the modeling language over time. Metamodel evolution may induce severe repercussions over the related artifacts, which might need to be consistently adapted. The co-evolution problem poses another issue related to the epistemic uncertainty arising as a response to a metamodel change where many different consistency restoration procedures are possible, e.g., restoring models and transformations. In this work we rely on the notion of information loss (IL) to understand which migration combinations are preferable to others in coupled evolution, by offering a ranking of the possible solutions. The IL denotes that part of the knowledge contained in the source models cannot be conveyed or translated into the target models, using a selected transformation, resulting in a loss of information. Such aspect represents an extra-functional quality that can be used for ranking the selection of certain transformation migrations in the context of multiple available alternatives and in combination with model migrations. Information loss induced by a (migrated) transformation can be defined as the amount of information lost executing the (migrated) transformation compared to the original output model obtained by executing the original transformation. The proposed approach is supported by a prototype tool that we demonstrate and validate.
2021
MDE , Parallel Coupled-evolution , Information loss.
File in questo prodotto:
File Dimensione Formato  
2021_JObjectTechnol_20_Basciani.pdf

accesso aperto

Tipologia: Versione Editoriale (PDF)
Licenza: Dominio pubblico
Dimensione 3.21 MB
Formato Adobe PDF
3.21 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12571/24366
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? 1
social impact