Software systems are becoming an integral part of every part of our life and as a result more software systems are produced than ever before. To manage the complexity and improve the efficiency of producing large varieties of software products, software product lines have received increasing support and attention. In a product line, the production of a variety of products is simplified by capturing the common properties of the products in reusable building blocks. Specific products are realized by extending the functionality of the blocks using sophisticated variation mechanisms. While product lines offer an effective approach to handle product families, their design is a very complex activity, which requires a considerable amount of information. Accurate information on the variance in products is required for the product line to be effective. While it is difficult enough to acquire such information under ideal circumstances, in practice software architects are faced with partial and ambiguous descriptions of the desired functionality and variability. This type of information is referred to as imperfect information. We address this problem by coming to an analysis approach and supporting tool that offer decision support during the development of product lines. The approach and tool assist the software developer in identifying imperfection in requirements and considering its impact on product line design. To achieve this objective, we utilize a multi-disciplinary approach combining techniques from corpus linguistics, fuzzy logic, aspect-oriented requirement analysis and combinatorial optimization. The research project offers excellent potential to further the excellence of the European Union in product line research and imperfect information in software design. With the state-of-the-art research topic, a comprehensive training program and the facilities offered by the host, the project offers outstanding potential for the development of the fellow’s career.
Fields of science
Call for proposal
See other projects for this call