1 of 18

1
CSA3080Adaptive Hypertext Systems I
Lecture 7Formal Models of HypertextDexter
Hypertext Reference Model

• Dr. Christopher Staff
• Department of Computer Science AI
• University of Malta

2
Aims and Objectives

• Adaptive Hypertext Systems are built using
  hypertext navigation support and information is
  inter-linked, hypertext style
• Most AHSs are deployed over the Web, but the Web
  isnt a particularly good example of a hypertext
• So what are the properties and characteristics of
  good hypertexts?

3
Aims and Objectives of Hypertext

• Well, by hypertext I mean non-sequential
  writing--text that branches and allows choices to
  the reader, best read on an interactive screen
  Ted Nelson, 1987. Literary Machines, Edition
  87.1.
• Hypertext is text which is not constrained to be
  linear. Hypertext is text which contains links to
  other texts. http//www.w3.org/WhatIs.html
• References
• http//www.google.com/search?qdefineHypertext

4
Hypertext 1988 and beyond

• The WWW was first launched in 1991, but only
  gained popularity in 1993
• The Hypertext community came to a head in 1990 to
  iron out many inconsistencies and
  incompatibilities in terminology
• Many models of hypertext were also proposed,
  based on petri-nets, sets, etc.
• The most popular model is based on graph theory

5
Dexter Hypertext Reference Model

• Why a formal model?
• The goal of the model is to provide a principled
  basis for comparing systems as well as for
  developing interchange and interoperability
  standards Halasz94
• DHRM has been implemented as Amsterdam, CMIFed,
  AHAM, DeVise/WebVise, RHYTHM (Bologna)
• References
• Halasz, F. and Schwartz, M. 1994. The Dexter
  Hypertext Reference Model, in Communications of
  the ACM, 37(2), February, 1994, 30-39.

6
DHRM Fundamentals

• DHRM separates the representation of documents
  (nodes) from the linking of nodes and the
  navigation through hyperspace

Halasz94
7
DHRM Fundamentals
8
DHRM Fundamentals

• Components
• DHRM components are the equivalent of nodes, and
  are represented in the Storage Layer
• Nodes were called frames, cards, documents, and
  articles
• Even today, on the Web a node is referred to as a
  document or more commonly, a page
• DHRM doesnt really care about what happens
  within a component, only how the hypertext
  network interfaces with the component

9
DHRM Fundamentals

• Anchors
• References to locations or items within documents
• Components can be composites, hierarchical
  combinations of atomic components (DAG)
• Anchors can be the source or destination of links
• Anchors can be entire components, or spans
  (segments of a component)

10
DHRM Fundamentals

• More about anchors
• Anchors have two parts
• Anchor ID (used by Storage Layer)
• Anchor Value (used by Within-Component Layer)
• The anchor value can be a region within a
  component, and the value is sensible only to the
  application responsible for editing/accessing the
  component
• Anchors are unique ltcomponentID, anchorIDgt

11
DHRM Fundamentals

• Links
• Links are represented in the Storage Layer
• Specify a source anchor, a destination anchor,
  and a direction that specifies how the link can
  be traversed
• Links can also be the destinations for other
  links
• Links, therefore, are totally separate from the
  components that contain them

12
DHRM Fundamentals

• Presentation Layer
• A hypertext isnt much good if you cannot
  manipulate it and navigate through it
• From the Presentation Layer users can access,
  view, and manipulate the hypertext

13
DHRM Fundamentals

• All components (including links) have
  presentation specifications
• The Presentation Layer can also impose
  presentation specifications on the accessed links
  and components to capture user preferences, for
  instance

14
Referring to components

• Components have unique identifiers (UIDs) and
  component specifications
• Component specifications are essential, because a
  user may be able to describe a component without
  knowing its UID
• Components may be identified from their
  description using a resolver function, and then
  retrieved using an accessor function

15
DHRM Fundamentals
Halasz94
16
DHRM Fundamentals

• More about links
• Links are first class objects
• Links are created by combining a component
  specification, anchor ID, direction, and
  presentation information into a specifier
• Direction can be FROM, TO, BIDIRECT, NONE
• A link is a sequence of two or more specifiers,
  at least one of which must be TO or BIDIRECT

17
Conclusion

• Interesting features of DHRM
• Links are separate from documents containing them
• Anybody can be an author (link creator)
• Search capability is built into hypertext model
  (resolver function)
• Presentation specifications can change behaviour
  of component when displayed
• Links know their origin and destination
• Components can be composite
• Dangling links are not allowed (supposedly!)

18
Conclusion

• DHRM was defined in 1990
• Most existing hypertext systems were small scale,
  catering for individuals and small workgroups
• The Internet (using TCP/IP) had existed for 7
  years
• The WWW did not yet exist