Service-based applications are typically composedof web services, which are selected from a pool of servicesand/or cloud market. Selection and composition decisions tendto encounter numerous uncertainties: service consumers andapplications have little control of these services and tend to beuncertain about their level of support for the desired functionalitiesand non-functionalities. We contribute to an "intelligent"framework for selecting and composing services. The frameworkis ground on the premise of computationally self-awareness toinform decisions of selecting and composing the services to meetboth behavioral and functional requirements. The frameworkprovides the primitives for fine grained representation of knowledgeand levels of self-awareness for time, goal, interaction andstimuli. We have used volunteer service computing as an exampleto demonstrate the benefits that self-awareness can introduce toself-adaptation.