| 摘要 |
<p>To carry out the method, an interpretation device, preferably computer-controlled, is provided, which contains a data base with stored document models and at least one main memory, into which, following recording, pre-processing and grammatical analysis using known methods, a document to be interpreted is read in as a document graphic. In order to identify the type of the form read in as a document graphic (D) and the meaning of the user entries on the form (F), various models (M) of different forms (F) stored in a knowledge base are mapped successively onto the symbolic description of the document in the form of a document graphic (D). <IMAGE></p> |
