Refactoring a legacy monolithic application into microservices involves discovering the current application’s architecture, refactoring it into a microservices architecture, and then restructuring the application based on the new architecture. This process is time-consuming and typically requires expert involvement. In this paper, we propose an approach that recommends candidate microservices from a given monolith, aligning them with business services for the architect. Our approach begins by recovering the architecture of the monolithic application using static analysis, then identifies clusters of cohesive application fragments through unsupervised learning. These clusters are aligned with business services as specified by a domain model. We have implemented our approach as a tool and compared its results with state-of-the-art approaches using a suite of four Java applications. The recommended clusters by our tool are highly cohesive, loosely coupled, and better aligned with domain services.