With XML as with SGML, we can have structural correctness, once they provide syntactic rules to state how to mark-up all the documents of the same family (of the same type); moreover, XML also imposes a working approach in which there is a complete separation between the structure of the data and the way it looks. So it is possible to avoid that someone will write a letter putting the ending before the body. Also, being purely declarative and completely independent of the processing, it is possible to swap documents between different systems without having to change them. But even this way, there still is a lack of content validation. Therefore, as Ramalho et al. pointed out, if a document has the decrees published by some kings, and includes their birth dates, it is critical if there is a sentence in which a king is said to publish a decree before he was born. In this paper we are concerned about reaching a way to automatically process a document in order to validate it semantically, avoiding this kind of incongruences that may spoil all a teams work.