We aim at automating an Integration process for Component-based Systems. Hence we describe a reference model where its first phase (Qualification) is our current work. We intend to compare components data by semantic analysis — i.e. Interoperability at a Semantic level. Assessment is based on meta-data added to components: their context, assertions, and interface protocol. In a previous work we showed a component interface matching. In this paper we explain the Assertions matching, that uses a technique based on Abstract Syntax Trees, which detects similar pieces of code (?clones?). We aim to enhance the procedure by a semantic component treatment using Ontologies. Thus the whole Assessment procedure is highly improved both on efficacy and reliability.