Software reuse is an important approach to increase software quality and productivity. In a software reuse procedure, extracting the software components with high reuse potential and high quality is a key step that may directly affect the steps that follow. Some software metrics can measure the quality characteristics of the RSC (Reusable Software Component). However, individual software metric cannot measure the overall quality characteristic of the RSC. Therefore, the software metrics shall be combined, and the conflict situations in metric combination should be reduced. In this paper, we propose a rule-based metric combination model that is based on dynamically weighted linear combinations and conflicts reducing production rules. Applying this model, a highly flexible and practical metric combination approach can be created, the conflict situations in metric combination can be reduced, and a RSC extraction tool can be generated.