Abstract
Ontology-driven content-based systems are content-based systems (ODCBS) that are built to provide a better access to information by semantically annotating the content using ontologies. Such systems contain ontology layer, annotation layer and content layer. These layers contain semantically interrelated and interdependent entities. Thus, a change in one layer causes many unseen and undesired changes and impacts that propagate to other entities. Before any change is implemented in the ODCBS, it is crucial to understand the impacts of the change on other ODCBS entities. However, without getting these dependent entities, to which the change propagates, it is difficult to understand and analyze the impacts of the requested changes. In this paper we formally identify and define relevant dependencies, formalizing them and present a dependency analysis algorithm. The output of the dependency analysis serves as an essential input for change impact analysis process that ensures the desired evolution of the ODCBS.