The software product line (SPL) is a promising approach to developing software systems that are complex, large-scale, and rich in product configuration. The automatic acquisition of product configurations’ models, which may be too many in the software product line, is very important in terms of time and cost constraints. In this study, we propose an approach that automatically extracts models of different product configurations from the full product model by feature deduction to obtain product models more productively and effectively. We validated the proposed approach on three different case studies: Soda Vending Machine SPL, Bank Account SPL, and Student Attendance System SPL. The two novelties of this study are the feature-dependency tree and the dynamic edge mapping algorithm.