Managing software product line using an ontological rule-based framework

Abstract

Feature model is amongst the widely used approach to manage variability and commonality in software product lines for enabling software reuse. Modeling variability and developing variable software products from feature model is an arduous task. Due to the growing complexity and size of feature models, defects such as false-optional and dead features can arise which lead to the development of low-quality erroneous software products. We propose an ontological rule-based framework that improves software product line by identifying feature model defects due to false-optional and dead features with their causes. This information further suggest corrections to the modelers for resolving defects. The evaluation results show the applicability of proposed framework to handle false-optional and dead features, and indicate its accuracy and scalability for huge size feature models with 21, 050 features. Thereby, improving the quality and reusability of software product line.