The huge number of available documents on the Web makes finding relevant ones challenging. Thus, searching for information becomes more and more complex because of the growing volume of data and of its lack of structure. The quality of results that traditional full-text search engines provide is still not optimal for many types of user queries. Especially, the ambiguities of natural languages and abstract concepts are handled inadequately by full-text search engines. Ontologies provide a solution to these problems. They can help a user to find documents of a specific domain. This paper proposes a new retrieval system based on ontologies. This system integrates results from traditional full-text engines, and thus supports a gradual transition from classical full-text search engines to ontology-based ones.