Modelling Imperfect Product Line Requirements with Fuzzy Feature Diagrams

In this article, we identify that partial, vague and conflicting information can severely limit the effectiveness of approaches that derive feature trees from textual requirement specifications. We examine the impact such imperfect information has on feature tree extraction and we propose the use of fuzzy feature diagrams to address the problem more effectively. This approach is then discussed as part of a general research agenda for supporting imperfect information in models that are used during variability analysis.