Ontologies published on the Semantic Web are becoming an increasingly popular means for sharing information and knowledge in various branches of science and economy. Ontologies, expressed in the standard Semantic Web languages like RDF or OWL, are getting bigger and bigger. Some of them, in particular ontologies in the medicine domain, are composed of thousands of concepts, roles and properties, which makes them hard to comprehend and manage. Hence there is a strong need for modularization of ontologies. Syntactic methods, such as the import mechanism used in OWL, have no underlying semantics, which makes them artificial and somewhat arbitrary. The lecture will present a new, semantic method of ontology modularization called Structural Interpretation Model (SIM). According to this method, an ontology is divided into modules based on values of contextual parameters. SIM is a special case of a more general mechanism for ontology modularization called tarsets. Tarsets are basic ontological modules that are analogous to relations in relational databases. The lecture will present examples of simple SIM-based ontologies, and illustrate the flow of conclusions between ontological modules, performed by means of expressions of tarset algebra.