Citation
Bernhard Rumpe. "Advanced topics in model-based software development". Talk or presentation, 3, June, 2008.

Abstract
Model-based software & systems development has not yet become that successful and efficient as the concept of modelling has promised. We are going to investigate, what models are, where the can be successfully used and what the obstacles of model-based developments are. We investigate in specific topics, such as formal semantics of models, model composition and refinement and model-based architectural evolution to understand what potential lies in the use of models. E.g. evolutionary approaches become increasingly important due to faster changes of requirements as well as technologies both in the business and in the embedded systems domain. We therefore show how software evolution can be managed on an architectural level provided that reliability ensuring mechanisms, such as test models exist. The key idea of the approach is to use model artifacts to describe the architecture of a system and others to model tests. An architecture is evolved using systematic refactoring techniques and thus becomes a lot easier when regression tests allow to repeatedly check the correctness of each evolution step and thus retain reliability of the product during the evolutionary process.

Electronic downloads

• 08.06.ASD.SanFrancisco.pdf · application/pdf · 268 kbytes
• rumpechess.pdf · application/pdf · 268 kbytes

• HTML

  Bernhard Rumpe. <a
  href="http://chess.eecs.berkeley.edu/pubs/433.html"
  ><i>Advanced topics in model-based software
  development</i></a>, Talk or presentation,  3,
  June, 2008.

• Plain text

  Bernhard Rumpe. "Advanced topics in model-based
  software development". Talk or presentation,  3, June,
  2008.

• BibTeX

  @presentation{Rumpe08_AdvancedTopicsInModelbasedSoftwareDevelopment,
      author = {Bernhard Rumpe},
      title = {Advanced topics in model-based software development},
      day = {3},
      month = {June},
      year = {2008},
      abstract = {Model-based software \& systems development has not
                yet become that successful and efficient as the
                concept of modelling has promised. We are going to
                investigate, what models are, where the can be
                successfully used and what the obstacles of
                model-based developments are. We investigate in
                specific topics, such as formal semantics of
                models, model composition and refinement and
                model-based architectural evolution to understand
                what potential lies in the use of models. E.g.
                evolutionary approaches become increasingly
                important due to faster changes of requirements as
                well as technologies both in the business and in
                the embedded systems domain. We therefore show how
                software evolution can be managed on an
                architectural level provided that reliability
                ensuring mechanisms, such as test models exist.
                The key idea of the approach is to use model
                artifacts to describe the architecture of a system
                and others to model tests. An architecture is
                evolved using systematic refactoring techniques
                and thus becomes a lot easier when regression
                tests allow to repeatedly check the correctness of
                each evolution step and thus retain reliability of
                the product during the evolutionary process.},
      URL = {http://chess.eecs.berkeley.edu/pubs/433.html}
  }
  

Posted by Hiren Patel on 13 Jun 2008.
Groups: chess
For additional information, see the Publications FAQ or contact webmaster at chess eecs berkeley edu.