Bachelor and Master Theses

Title: Horizontal transformations for models reuse and tool chaining
Subject: Computer Science
Level: Advanced
Description: The increasing attention of industry on model-driven techniques pushed tool vendors toward the design and release of a new generation of utilities able to support such methodology. Because of the need to keep the same quality factors if compared to the existing environments, very often providers realized tools tailored to a specific applicative domain, and for a specific task. On the other hand, model-driven development is a vision which crosscuts the whole software lifecycle, making the exploitation of model-driven techniques spread among the development process. As a consequence, the industries are facing an increasing need to build up a shared model repository by which could be possible to reuse the same models both for design and testing, for instance. Unfortunately, as clarified above tailored solutions hamper this demand due to different formats and type of information requested to the user. For example, the same version of the UML 2 metamodel can be interpreted in disparate ways, depending on the kind of use that is done of the produced models. Therefore, a system model produced for design purposes cannot be reused for testing and validation, but requires the developers to provide new models adapted to the latter scope. It is worth noting that such duplication is not only a time consuming activity, but can also be error prone and introduce inconsistencies between the various artefacts.

In model-driven engineering model transformations play the fundamental role of mapping engines able to derive target models from source models, as well as other useful artefacts like documentation and code. Therefore, this work investigates the chance to support the exchange of models between tools through the use of model transformations. In particular, a pivot language can be described as the single source of information by which all the other formats can be (partially) derived. Obviously, since the pivot language has to be necessarily generic, it will abstract some specific information tailored to the particular kind of use the model is devoted to. In any case, it enables the set up of a coherent skeleton, with respect to the shared definition of the system, in which the developer can input those tool specific data. Moreover, if for some reason one of the tool formats is revised, it is sufficient to update the mapping between such tool format and the pivot model, while all the other mappings are preserved, thus also preserving the automatic generation.
Company: Ericsson, kontaktperson: Pär Asplund
Prel. end date: 2010-12-15
Presentation date: 2011-02-03
Student: Stefano Cucchiella
IDT supervisor: Antonio Cicchetti, +46-21-151762
Examinator: Mikael Sjödin
Mikael Sjödin, +46 70 288 2829
Misc: Work plan: - study of BridgePoint and RSA formats through industrial case studies; - definition of a prototypical pivot language in the Papyrus open source environment; - definition of mapping engines between BridgePoint, RSA, and the pivot; - valid

Rapport och bilagor


Senaste uppdatering

  • Mälardalen University |
  • Box 883 |
  • 721 23 Västerås/Eskilstuna |
  • 021-101300, 016-153600 |
  • webmaster |
  • Latest update: 2018.08.31