Applying Multi-Level Modeling to DataIntegration in Product Line Engineering
2017 (English)In: CEUR Workshop Proceedings: MODELS 2017 Satellite Events, CEUR Workshop Proceedings , 2017, Vol. 2019, p. 235-242Conference paper, Published paper (Refereed)
Abstract [en]
Developing systems according to the Product Line Engineer-ing (PLE) paradigm is a process in which dierent types of engineeringartifacts are created with the aim of reusing them in dierent congurationsof the same system. Ensuring that dierent system congurationssatisfy various functional and non-functional properties is ensured byanalyzing dierent artifacts but because they are maintained by dierenttools, sometimes even manually, achieving and especially automatingsuch analyses is a challenging task. Overcoming this issue can beachieved through data integration of existing data which implies creatingan information model that species how will the existing data fragmentsbe related, captures relevant domain constraints, and most importantlycaptures the fact that some data objects are classes in one tool and instancesin another. This paper reports on the experiences from applyingthe Multi-Level conceptual Theory (MLT), to the problem of informationmodeling for data integration in the PLE context. Being a Multi-LevelModeling, powertype-based framework, MLT allows separation of the classand instance facet of modeled entities while keeping them explicit. Someof the MLT modeling constructs are particularly useful for capturing therenement levels of the modeled artifacts and for succinctly capturingconstraints like disjointess or completeness among them. We also reportthat there were certain aspects of the studied case that could not be expressedusing MLT. The studied case comes from a real data-integrationproject from the heavy vehicle manufacturer, Scania CV AB.
Place, publisher, year, edition, pages
CEUR Workshop Proceedings , 2017. Vol. 2019, p. 235-242
Series
CEUR Workshop Proceedings, ISSN 1613-0073 ; 2019
Keywords [en]
Multi Level Modeling, Information Modeling, Product Line Engineering, Linked Data
National Category
Computer Sciences
Research subject
Computer Science
Identifiers
URN: urn:nbn:se:kth:diva-220731Scopus ID: 2-s2.0-85041435512OAI: oai:DiVA.org:kth-220731DiVA, id: diva2:1170822
Conference
2017 MODELS Satellite Event: Workshops (ModComp, ME, EXE, COMMitMDE, MRT, MULTI, GEMOC, MoDeVVa, MDETools, FlexMDE, MDEbug), Posters, Doctoral Symposium, Educator Symposium, ACM Student Research Competition, and Tools and Demonstrations, MODEL-SSE 2017, Austin, United States, 17 September 2017 through 22 September 2017
Projects
ASSUME
Note
QC 20180123
2018-01-042018-01-042018-03-16Bibliographically approved