THE USE OF A FUNCTIONAL GRAMMAR IN THE AUTOMATIC PROCESSING OF STYLISED ADMINISTRATIVE DOCUMENTS IN A LAND REGISTRY

This paper presents an approach inspired by the Functional Grammar (fg) of S.C. Dik, which has been rethought to adapt it to the context of automatic analysis of bureaucratic documents. The research is part of a larger ongoing project for the automated processing of land registry documents and the model presented has been demonstrated for automatic content analysis of documents concerning technical operations of real estate (division and regrouping of lots) and is to be extended also to mortgages. The first paragraphs examine the theoretical principles of the application; the process of understanding the document content through the production of predications is then presented and the content analysis strategy briefly sketched.