Most of the existing textual information retrieval approaches depend on a lexical match between words in user's requests and words in target objects. Typically only objects that contain one or more common words with those in the user's query are returned as relevant. This lexical based retrieval model is far from ideal. In this research an approach to semantic based information retrieval of semantically annotated documents is presented. The approach operates based on: (i). natural language understanding, (ii). the Wordnet ontology, and (iii). the Semantic web standards. Not only the information is annotated and searched on a semantic basis, but also the retrieval process can be enhanced by the use of rich vocabulary knowledge in the ontology.