Database Support for Very Large Hypertexts

Current hypertext systems have been widely and effectively used on relatively small data volumes. The potential of database technology is explored for aiding the implementation of hypertext systems holding very large amounts of complex data. Databases meet many requirements of the hypermedium: persistent data management, large volumes, data modelling, multi-level architecture with abstractions and views, metadata integrated with operational data, short-term transaction processing and high-level end-user languages for searching and updating data. To illustrate the potential for the use of databases, a system implementing the storage, retrieval and recall of trails through hypertext comprising textual complex objects is described. Weaknesses in current database systems for handling the complex modelling required are discussed.