This paper presents a quantified approach to selecting architectural tactics using architectural tactic knowledge base. An architectural tactic is an architectural design building block pertaining to a specific software quality. The tactic knowledge base consists of structural and behavioral architectural tactic semantics specified in Role Based Metamodeling Language(RBML) and their inter-relationships expressed in feature model. In this paper, a cost of an architectural tactic is estimated with use case point, while relative significance of non-functional requirements are predicted by using AHP(Analytic Hierarchy Process) and sensitivity analysis method. Base on it, our approach suggests the best possible fit which is likely to satisfy non-functional requirements. We applied the approach to build software architecture of an on-line trading system.