You are here: Home / Publications

Scientific Papers

ISSI Research Papers

Paper information
Title
A Comparative Study on Model-Driven Requirements Engineering for Software Product Lines
Published in
Revista de Sistemas e Computação 2(1), ISSN 2237-2903, Universidade Salvador (Brazil) - 2012
Abstract
Model-Driven Engineering (MDE) and Software Product Lines (SPL) are two software development paradigms that emphasize reusing. The former reuse domain knowledge is represented as models and model transformations for product development, and the latter reuse domain knowledge is represented as core assets to produce a family of products in a given domain. The adequate combination of both paradigms can bring together important advantages to the software development community. However, how to manage requirements during a model-driven product line development remains an open challenge. In particular, the Requirements Engineering (RE) activity must deal with specific properties such as variability and commonality for a whole family of products. This paper presents a comparative study of eleven approaches that perform a MDE strategy in the RE activity for SPL, with the aim of identifying current practices and research gaps. In summary, most of the approaches are focused on the Domain Engineering phase of the SPL development, giving less attention to the Application Engineering phase. Moreover there is a lack of coverage of the Scoping activity, which defines the SPL boundaries. Several approaches apply some model transformations to obtain architectural and application requirements artifacts. Regarding the tool support for requirements specification and management, we found that most of the approaches use only academic prototypes. Regarding the validation of the approaches, the use of Case Studies as a proof of concept was the most commonly used method; however, there is a lack of well-defined case studies and empirical studies to improve the proposals.


BibTeX
@misc{issi_web:id:412,
        title =  "A Comparative Study on Model-Driven Requirements Engineering for  Software Product Lines ",
        author = "David Blanes Domínguez and Emilio Insfrán Pelozo",
        booktitle = "Revista de Sistemas e Computação 2(1), ISSN 2237-2903, Universidade Salvador (Brazil)",
        year = "2012",
        eprint = "http://issi.dsic.upv.es/publications/archives/",
        url = "http://www.revistas.unifacs.br/index.php/rsc/article/view/2174",
        abstract = "Model-Driven Engineering (MDE) and Software Product Lines (SPL) are two software development paradigms that emphasize reusing. The former reuse domain knowledge is represented as models and model transformations for product development, and the latter reuse domain knowledge is represented as core assets to produce a family of products in a given domain. The adequate combination of both paradigms can bring together important advantages to the software development community. However, how to manage requirements during a model-driven product line development remains an open challenge. In particular, the Requirements Engineering (RE) activity must deal with specific properties such as variability and commonality for a whole family of products. This paper presents a comparative study of eleven approaches that perform a MDE strategy in the RE activity for SPL, with the aim of identifying current practices and research gaps. In summary, most of the approaches are focused on the Domain Engineering phase of the SPL development, giving less attention to the Application Engineering phase. Moreover there is a lack of coverage of the Scoping activity, which defines the SPL boundaries. Several approaches apply some model transformations to obtain architectural and application requirements artifacts. Regarding the tool support for requirements specification and management, we found that most of the approaches use only academic prototypes. Regarding the validation of the approaches, the use of Case Studies as a proof of concept was the most commonly used method; however, there is a lack of well-defined case studies and empirical studies to improve the proposals."
}