A two‐level structure for textual databases to support hypertext access

This article introduces a model of textual databases in which the structure of the text is divided into two components: the syntactic structure and the access structure. The syntactic structure of the text is defined by a schema, which is context‐free grammar. The contents of a text database defined by such a schema is a parse tree for the schema. The schema defines the hierarchical relationships of the text parts to each other. The access structure, which is defined during access to the data by the user, is modeled on a hypergraph, generated from the syntactic structure by the user's queries. The queries are expressed by constrained productions. In browsing, the hypergraph will be viewed as a graph. We will show that this two‐level model for text can be used to extend hypertext access capabilities to text databases that were not originally authored to be hypertext. © 1992 John Wiley & Sons, Inc.