Abstract The enormous number of data stored in large, relational databases makes the exploring, querying and reporting processes difficult. The database queries executed on redundant, incoherent, and contextless data may cause erroneous results. The Semantic Web (SW) technology has been considered to solve the problem. In this paper we propose ControlSem - the semantic oriented system architecture for controlling medical procedures. We performed the set of inference tasks that confirmed the usefulness of the SW technology. The experience with ControlSem modeling can be used for implementation of solutions to the contextual data integration and analysis.