In the field of software architecture, there has been a paradigm shift from describing the outcome of architecting process to documenting Architectural Knowledge (AK), such as design decisions and rationale. To this end, a series of domain models have been proposed for defining the concepts and their relationships in the field of AK. To a large extent, the merit of this new paradigm is derived by sharing and reusing AK across organizations, especially in geographically distributed contexts. However, the employment of different AK domain models by different parties makes effective AK sharing challenging, as it needs to be mapped either from one domain model to another directly, or indirectly through a central model for simplicity when the number of AK models increases. The indirect mapping approach has proved to be a cost-effective way by sacrificing acceptable sharing quality compared with direct mapping approach. However, there exist no criteria for the selection of a high quality central model besides the intuitive judgment by domain experts. In this paper, we propose to tackle this issue by using the concept of semantic distance between AK models, which is calculated using rules based on the concept mapping relationships between the models. A high quality central model is therefore the one with the shortest semantic distance to all potential AK models.