Open AccessFeature Model Configuration Based on Two-Layer Modelling in Software Product Lines
Author(s) -
Elham Farahani,
Jafar Habibi
Publication year - 2019
Publication title -
international journal of electrical and computer engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.277
H-Index - 22
ISSN - 2088-8708
DOI - 10.11591/ijece.v9i4.pp2648-2658
Subject(s) - computer science , software product line , software , context (archaeology) , process (computing) , feature model , software requirements , layer (electronics) , feature (linguistics) , software engineering , product (mathematics) , functional requirement , user requirements document , software development , systems engineering , software construction , operating system , engineering , paleontology , linguistics , chemistry , philosophy , geometry , mathematics , organic chemistry , biology
The aim of the Software Product Line (SPL) approach is to improve the software development process by producing software products that match the stakeholders’ requirements. One of the important topics in SPLs is the feature model (FM) configuration process. The purpose of configuration here is to select and remove specific features from the FM in order to produce the required software product. At the same time, detection of differences between application’s requirements and the available capabilities of the implementation platform is a major concern of application requirements engineering. It is possible that the implementation of the selected features of FM needs certain software and hardware infrastructures such as database, operating system and hardware that cannot be made available by stakeholders. We address the FM configuration problem by proposing a method, which employs a two-layer FM comprising the application and infrastructure layers. We also show this method in the context of a case study in the SPL of a sample E-Shop website. The results demonstrate that this method can support both functional and non-functional requirements and can solve the problems arising from lack of attention to implementation requirements in SPL FM selection phase.