With the advent of Cloud Computing, new opportunities for Business Process Outsourcing services have emerged. Business Process as a Service (BPaaS), a new cloud service model, has recently gained a great importance for outsourcing cloud-based business processes constructed for multi-tenancy. In such a multi-tenant environment, using configurable business process models enables the sharing of a reference process among different tenants that can be customized according to specific needs. With a large choice of configurable process modeling languages, different providers may deliver configurable processes with common functionalities but different representations which makes the process discovery and configuration a manual tedious task. This in turn creates cloud silos and vendors lock-in with non-reusable configurable BPaaS models. Therefore, with the aim of enabling the interoperability between multiple BPaaS providers, we propose in this paper a semantic framework for BPaaS configurable models. Taking advantage of Semantic Web technologies and data mining techniques, our framework allows for (1) an ontology-based high level abstract representation of BPaaS configurable models enriched with configuration guidelines and (2) an automated approach for extracting the configuration guidelines from existing process repositories. To show the feasibility and effectiveness of our approach, we extend Signavio with our semantic framework and conduct experiments on a dataset from SAP reference model