Representing and Utilising Knowledge for Understanding Structured Documents

This paper presents a document analysis system which is capable of extracting the semantics of specific text portions of structured documents. The main component of the system is the knowledge representation scheme - called h s c o , Frame Representation of Structured Documents. It allows the definition of knowledge about document components as well as knowledge about analysis algorithms in a uniform, simple, but powerful representation formalism. The specificinference algorithm of Frescois presented which combines the two different knowledge sources -the document model and the algorithmic model. The flexibility of the representation formalism Fresco and the properties of the inference algorithm are shown in two different applications, in interpreting amount fields on cheques and in analysing business letters.

• Type: Conference
• Year: 1992
• Where: MVA
• Authors: Thomas Bayer
• Keywords: Knowledge Representation Scheme, MVA 1992, MVA 2007, Representation Formalism, Structured Documents