Faezeh Ensan Publication Page

Dynamic Decision Models for Staged Software Product Line Configuration

Ebrahim Bagheri and Faezeh Ensan
Reference:
Ebrahim Bagheri and Faezeh Ensan Dynamic Decision Models for Staged Software Product Line Configuration. In Requirements Engineering Journal, 19 (2): 187-212, 2014.
Links to Publication: [doi]
Abstract:
Software product line engineering practices offer desirable characteristics such as rapid product development, reduced time-to-market and more affordable development costs as a result of systematic representation of the variabilities of a domain of discourse that leads to methodical reuse of software assets. The development lifecycle of a product line consists of two main phases: domain engineering, which deals with the understanding and formally modeling of the target domain; and application engineering that is concerned with the configuration of a product line into one concrete product based on the preferences and requirements of the stakeholders. The work presented in this paper focuses on the application engineering phase and builds both the theoretical and technological tools to assist the stakeholders in (1) understanding the complex interactions of the features of a product line; (2) eliciting the utility of each feature for the stakeholders and hence exposing the stakeholders' otherwise implicit preferences in a way that they can more easily make decisions; and (3) dynamically building a decision model through interaction with the stakeholders and by considering the structural characteristics of software product line feature models, which will guide the stakeholders through the product configuration process. Initial exploratory empirical experiments that we have performed show that our proposed approach for helping stakeholders understand their feature preferences and its associated staged feature model configuration process is able to positively impact the quality of the end results of the application engineering process within the context of the limited number of participants. In addition, it has been observed that the offered tooling support is able to ease the staged feature model configuration process.
Bibtex Entry:
@article{DBLP:journals/REJ/BagheriE13, author = {Ebrahim Bagheri and Faezeh Ensan}, title = {Dynamic Decision Models for Staged Software Product Line Configuration}, journal = {Requirements Engineering Journal}, volume = {19}, number = {2}, year = {2014}, pages = {187-212}, ee = {http://link.springer.com/article/10.1007%2Fs00766-013-0165-8}, abstract = {Software product line engineering practices offer desirable characteristics such as rapid product development, reduced time-to-market and more affordable development costs as a result of systematic representation of the variabilities of a domain of discourse that leads to methodical reuse of software assets. The development lifecycle of a product line consists of two main phases: domain engineering, which deals with the understanding and formally modeling of the target domain; and application engineering that is concerned with the configuration of a product line into one concrete product based on the preferences and requirements of the stakeholders. The work presented in this paper focuses on the application engineering phase and builds both the theoretical and technological tools to assist the stakeholders in (1) understanding the complex interactions of the features of a product line; (2) eliciting the utility of each feature for the stakeholders and hence exposing the stakeholders' otherwise implicit preferences in a way that they can more easily make decisions; and (3) dynamically building a decision model through interaction with the stakeholders and by considering the structural characteristics of software product line feature models, which will guide the stakeholders through the product configuration process. Initial exploratory empirical experiments that we have performed show that our proposed approach for helping stakeholders understand their feature preferences and its associated staged feature model configuration process is able to positively impact the quality of the end results of the application engineering process within the context of the limited number of participants. In addition, it has been observed that the offered tooling support is able to ease the staged feature model configuration process. } }




Powered by WordPress