Detecting and Explaining Conflicts in Attributed Feature Models
This work addresses errors in variability models for product configuration systems, offering a domain-specific solution to improve model maintenance and reliability.
The paper tackles the problem of detecting and explaining contradictions in attributed feature models, which arise during product configuration system development, by proposing an approach that translates models into constraint satisfaction problems and uses the QuickXplain algorithm to identify conflicting relations, achieving efficient assistance for developers in correcting mistakes.
Product configuration systems are often based on a variability model. The development of a variability model is a time consuming and error-prone process. Considering the ongoing development of products, the variability model has to be adapted frequently. These changes often lead to mistakes, such that some products cannot be derived from the model anymore, that undesired products are derivable or that there are contradictions in the variability model. In this paper, we propose an approach to discover and to explain contradictions in attributed feature models efficiently in order to assist the developer with the correction of mistakes. We use extended feature models with attributes and arithmetic constraints, translate them into a constraint satisfaction problem and explore those for contradictions. When a contradiction is found, the constraints are searched for a set of contradicting relations by the QuickXplain algorithm.