Most of the thrust in the semantic web movement comes from the observation that existing NLP tools are not sophisticated or efficient enough to process the full richness of Natural Language, and therefore Machine Understandable annotations need to be added to Web Resources in order to make them accessible by remote agents. However, when the target application is not required to handle a huge amount of documents, but more limited sets, it is conceivable and practical to take advantage of NLP tools to pre-process textual documents in order to generate annotations (to be verified by human editors).
We discuss an approach based on a combination of various Natural Language Processing techniques that addresses this issue. Documents are analized fully automatically and converted into a semantic annotation, which can then be stored together with the original documents. It is this annotation that constitutes the machine understandable resource that remote agents can query.