HyperTeX — A System for the Automatic Generation of Hypertext Textbooks from Linear Texts

Abstract

This paper reports on the HyperTeX system from which electronic textbooks have been produced. In order to reconstruct the implicit, network-like structure of a textbook, basic ideas of the hypertext concept were used to develop the system. HyperTeX consists of two parts: One component (HyperTeX/G) generates a hypertext structure from a textbooks’s linear, machine-readable version and stores this strucutre in a relational database system, whereas the other component (HyperTex/B) provides for access to the hypertext textbook by means of a window- and mouse-oriented user interface.

HyperTeX/G’s task is on the one hand to recognize which units from the input text form hypertext nodes, on the other hand to discover explicit and implicit interrelations automatically and to establish them as hypertext links. Although the resulting hypertext has a static character at first, its links can be altered dynamically by drawing conclusions from instances of use of the system. HyperTeX/B has the special feature that teaching material can be displayed on a high resolution graphic screen in full photo type setting quality. Still, each word and parts of formulae are mouse-sensitive in the raster graphic.