Research on the Merging of Feature Models
Zhao Hai · Chinese Journal of Computers · 2013
Feature models provide an effective way to organize and reuse software requirements in a specific domain.Constructing a feature model needs a systematic analysis of as many applications as possible in a domain,to identify commonality,variability,and dependencies among requirements.With the increasing complexity of domains,the scale of feature models can be extremely large,and the construction of large feature models is an overwhelming task for human that computer-aided automation is needed.A feasible way is to merge existing feature models into a large one,and human developers only need to do some refactoring work.In this paper,we survey six methods of merging feature models.We propose a conceptual framework first,and then analyze and compare the six methods.Finally,we identify three problems in existing research,and propose possible ideas to handle these problems.