[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

[ba-ohs-talk] FOHM--the Fundamental Open Hypertext Model

ription.html (one of dozens  of pages <~1030> when you google FOHM)    (01)

"The Fundamental Open Hypertext Model (FOHM) [6] grew out of the Open 
Hypermedia Protocol (OHP) developed by the Open Hypermedia Systems Working 
Group (OHSWG) but it expands the OHP data model to describe a broader set 
of hypermedia "domains". FOHM also makes no assumptions about the protocol 
it is running over or the systems that are using it. It is a model for 
describing hypertext structures that requires binding to a syntax before it 
can be used"    (02)

"In its work on interoperability, the OHSWG considered the requirements of 
several domains of hypertext. The three most frequently mentioned were 
Navigational, Spatial and Taxonomic Hypertext. The OHP protocol was always 
more concerned with Navigational Hypertext, however FOHM is capable of 
expressing all three domains. Before we can examine FOHM it is necessary to 
define these domains.
Navigational Hypertext is the most traditional domain of hypertext, 
exemplified in Open Hypermedia Systems such as Chimera[1], DHM[3], 
HyperForm[11], Microcosm[2] and the HB/SP series[9]. Authors create Links 
between parts of documents that are related. Users can then click on those 
links to move between documents. Although Navigational Hypertext systems 
can be quite sophisticated, by far the most popular system, the World Wide 
Web, is also one of the simplest.
Spatial Hypertext systems allow users to organise their information 
visually in a process known as "Information Triage"[5]. Relationships 
between nodes are expressed by their visual characteristics such as 
proximity, colour or shape. This results in visual collections, or Spaces, 
such as lists and sets. Spatial hypertext systems are therefore ideal for 
an evolving organisation of data. Examples of such systems include VIKI[4] 
and CAOS[8].
Taxonomic Hypertext is the organisation of information artifacts into 
Categories[7]. Where authors disagree about the categorisation, the 
Taxonomy can branch into different Perspectives[10]. Applications can allow 
users to navigate the information space by moving between overlapping 
Categories and can also reason about the relationships that artifacts have 
with one another. "    (03)

"In FOHM we describe four first-class objects that are analogous to the 
objects in the OHP data model. Associations are structures that represent 
relationships between Data objects. These Data objects are wrappers for any 
piece of data that lies outside of the scope of the model. They normally 
represent a document although one could represent any file, stream or other 
item. It is exactly this feature that we shall use to make links between 
concepts and queries. Data objects are not directly placed in the 
Associations. Instead Reference objects are used, these either point at 
Data objects in their entirety or at parts of those Data objects, for 
example the second paragraph of a text document, or the second scene of a 
film. They are attached to the Association object via Bindings. Each 
Association also has a structure type and a feature space; each Binding 
must state its position in that feature space, effectively stating how it 
is bound to the Association's structure. "    (04)

"Where the FOHM model differs from other open hypermedia models is in its 
placement of context at the heart of the information. Context is left as an 
opaque object within the FOHM model with the specifics being defined by the 
Context objects can be attached to all parts of the FOHM structure ..."    (05)

"In addition to context, behaviour objects can also be attached to the 
different parts of the FOHM data structures. Like context objects, 
behaviour objects are also opaque. Unlike context objects however the 
implementation of the FOHM model is not required to understand the contents 
of the behaviour object..."    (06)

Originally, according to the Hypertext 2001 proceedings, there was an 
implementation called "Auld Leaky."  Now, it is known as "Auld Linky"
complete with downloads (it's Perl, I don't yet know the license).    (07)

What's particularly curious (for me) is the resemblance to topic maps.  It 
strikes me that their "contexts" are a topic map's "scopes", though topic 
maps don't do "behaviors."  FOHM appears to be association-centric, while 
topic maps appear to be topic-centric.  I'd really like to see what others 
think about FOHM.    (08)

Jack    (09)