A prototype electronic encyclopedia
ACM Transactions on Information Systems (TOIS)
The multimedia object presentation manager of MINOS: a symmetric approach
SIGMOD '86 Proceedings of the 1986 ACM SIGMOD international conference on Management of data
An object-oriented approach to multimedia databases
SIGMOD '86 Proceedings of the 1986 ACM SIGMOD international conference on Management of data
A system-controlled multi-type specialization hierarchy
Proceedings from the first international workshop on Expert database systems
Informational zooming: an interaction model for the graphical access to text knowledge bases
SIGIR '87 Proceedings of the 10th annual international ACM SIGIR conference on Research and development in information retrieval
Proceedings of the 1980 workshop on Data abstraction, databases and conceptual modeling
COLING '86 Proceedings of the 11th coference on Computational linguistics
Automatic generation of hypertext knowledge bases
COCS '88 Proceedings of the ACM SIGOIS and IEEECS TC-OA 1988 conference on Office information systems
On generalizing the concept of hypertext
MIS Quarterly
Dialogue strategies for multimedia retrieval: intertwining abductive reasoning and dialogue planning
MIRO'95 Proceedings of the Final conference on Multimedia Information Retrieval
Hi-index | 0.00 |
A common feature of various recently developed information systems is the decomposition of linear document structures which are enforced by conventional print media. Instead, a network organization of information units of different forms (textual, graphical, pictorial and even auditive presentation modes may be combined) is provided. Documents organized this way are called “hypertexts”. However, two questions arise immediately when an effort is made to build information systems on the basis of this conception:What are the “units” constituting a hypertext?What sort of links between the units will be provided?Most approaches to hypertext systems impose the task of deciding these questions on the authors of hypertexts, thus the systems are hypertext management devices (e.g. CHRISTODOULAKlS ET AL. 86, WOELK ET AL. 86). The approach taken in this paper leaves a more active role to the software by applying knowledge based techniques.The starting point is the automatic content analysis of machine-readable full-text documents which may be downloaded from a full-text data base. The analysis process results in a partitioning of the document into thematically coherent text passages, which are one kind of node of the hypertextual version of this document. Other nodes contain graphics, tables and summarizations. The content analysis is accomplished by a semantic parser, which has access to an explicit model of the discourse domain. The TOPIC-System (HAHN/REIMER 86) comprises prototypical implementations of these components. Due to the semantic modeling relations between the nodes may be formally defined in order to provide content oriented browsing facilities. The graphical retrieval system TOPOGRAPHIC (THIEL/HAMMWÖHNER 87) employs an already implemented subset of them to guide users to relevant text parts.In this paper we outline a structure model for hypertexts based on partial representations of the meaning of text parts. Formal definitions of content oriented relations between such text units are given in terms of a logic specification language.