Browse Prior Art Database

Automatic Hypermedia Link Generation

IP.com Disclosure Number: IPCOM000108630D
Original Publication Date: 1992-Jun-01
Included in the Prior Art Database: 2005-Mar-22
Document File: 2 page(s) / 86K

Publishing Venue

IBM

Related People

Baber, RL: AUTHOR

Abstract

Authoring of a hypermedia document requires extensive knowledge of available related documents. Large hypermedia systems have many possible links to this related information. These links may be hard coded or defined by the user, but in either case, the job may be immense for a large system. In addition, documents may reside in flat files, in various databases, or on CD-ROM. Solving such pragmatic issues greatly enhances the potential of hypermedia for meaningful use in the real world, for example, in on-line business, medical, and insurance multimedia reference databases, or replacement of voluminous software documentation with revisable "introductory" files distributed to each workstation.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Automatic Hypermedia Link Generation

       Authoring of a hypermedia document requires extensive
knowledge of available related documents.  Large hypermedia systems
have many possible links to this related information.  These links
may be hard coded or defined by the user, but in either case, the job
may be immense for a large system.  In addition, documents may reside
in flat files, in various databases, or on CD-ROM.  Solving such
pragmatic issues greatly enhances the potential of hypermedia for
meaningful use in the real world, for example, in on-line business,
medical, and insurance multimedia reference databases, or replacement
of voluminous software documentation with revisable "introductory"
files distributed to each workstation.  These small files would link
to server-based reference material, replacing the shelves of manuals
and adding video, audio, and image media to augment text.

      The solution is to automatically generate hypermedia links by
integrating an existing link management system with a semantic text
parser and database access systems.  These three major components are
"glued" together using two software components: the global search
engine and the discriminator.  A source document acts as a "seed" for
input to the AIDA semantic parser.  This tool generates four output
documents: a Table of Contents, an Index, a list of keywords, and a
Summary.  The list of keywords is input to the global search engine,
along with information as to the position of each word in the
original document.  This global search engine generates a query to
each of the available databases based on the subject of the keyword
in question. The discriminator component very simply keeps a list of
documents retrieved from the databases, forwarding each one both to
the link manager for link generation and back to AIDA for generation
of further links from that document. The Link Manager component may
be either the Link Manager component of [*] or...