Violatility analysis framework for product lines

Abstract

Evolution of a software intensive system is unavoidable. In fact, evolution can be seen as a part of reuse process. During the evolution of the software asset, the major part of the system functionality is normally reused. So the key issue is to identify the volatile parts of the domain requirements. Additionally, there is promise that tailored tool support may help supporting evolution in software intensive systems. In this paper, we describe the volatility analysis method for product lines. This highly practical method has been used in multiple domains and is able to express and estimate common types of evolutional characteristics. The method is able to represent volatility in multiple levels and has capacity to tie the volatility estimation to one product line member specification. We also briefly describe current tool support for the method. The main contribution of this paper is a volatility analysis framework that can be used to describe how requirements are estimated to evolve in the future. The method is based on the definition hierarchy framework.

Analyzing evolution of variability in a software product line: From contexts and requirements to features

“…By interpreting and recognizing the variation evolution trend, the understanding on how the SPL requirements have evolved can be improved greatly. By foreseeing variation evolution, an SPL can be instrumented with appropriate variability mechanisms accordingly [26] [25]. Thus, the time and cost needed in the reactive SPL refactoring and extensions can be reduced.…”

“…These works focus on the modeling and specification of volatile requirements only, and it is not explicit how to discover them. Savolainen et al [25] propose a volatility analysis framework to estimate evolution from two perspectives: the probability of a change in requirement in the future, and vagueness of the requirement. However, their work does not recognise the influences of context diversity in variability evolution.…”

Issue-based variability management

“…Aspect orientation [25] manages variability at the level of implementation models and source code but does not over requirements engineering. Approaches based on traceability [26], model fragmentation [27] and volatility analysis [28] structure modeling space to enable evolution of variability models. A rationale-based approach to support evolution of variability complements all the concepts mentioned above.…”

On the communication problem between domain engineering and application engineering

“…Existing product instantiation approaches based on augmentation of variability models [6], model-based configuration systems [7] and feature configurations [8] only focus on the part of product that could be derived automatically, but does not focus on conflicts-of-interests between domain and application engineers. Further, the existing product line evolution approaches based on aspect orientation [9], traceability [10] and volatility analysis [11] as well as product line scoping approach proposed by. Noor et al [12,13] does not support the communication problem between domain and application engineering.…”