Nodes & Links (I accepted Anchors and Navigation also, although not all systems have them). Not all systems actually use these terms (e.g.,the Dexter model doesn't use the word "node", but the object is there - called a component)
With its concept of layering.
Please check out: www.win.tue.nl/2L670/static/architecture.html
All multimedia models were influenced by the layer concept of OSI.
The differences are in the number and functionality of layers, whether multimedia is explicitly stated, what the objects and atomic elements are, and in the ways the basic elements (nodes, links, ...) are modeled.
The Dexter Model is primarily a Hypertext model. It has 3 layers with 2 interfaces. Users interact with the Runtime layer, while the major focus of the model is on the storage layer, which describes how the hypertext components (nodes) and limks are connected. The within-component layer is responsible for content selection via anchors. The Dexter Model was an attempt to capture the important abstractions found in past, and future hypertext systems.
The Amsterdam Hypermedia Model adds time and context to the Dexter model. Atomic elements in the Amsterdam Hypermedia Model are single assets. The Amsterdam Hypermedia Model supports multimedia composites and anchors can be associated with a time-span or interval of play.
Interactive Multimedia Association'smodel has a Meta Multimedia Data Definition layer.
The top, middle and lowest layers of hypertext architecture are, respectively, the Presentation Level, the Hypertext Abstract machine, and the database level.
HAM was the 1st hypertext system to define a reference model.
Trellis is based on Petri Nets, while other models use other objects.
A single asset
These next 3 questions were a problem for most people. I will probably be more explicit next year. Many people answered them for the Dexter model, rather than in general. Look at www.win.tue.nl/2L670/static/architecture.html
The Presentation layer
The hypertext abstract machine
The database
Yes, because of the implementation of anchors & component instantiation with bidirectional links. See: www.win.tue.nl/2L670/static/dexter-anchor.html and www.win.tue.nl/2L670/static/dexter-runtime.html There were some persuave arguments that the answer is no because of the static nature of this model, but I believe the answer is "yes"
No. Time and sychronization were not specified in this model. The persuasive argument for "yes" here revolved around the fact that the Dexter Model doesn't preclude describing aspects such as delayed image loading, so in this sense it can be "expressed". I'd need more convincing for this.
Artificial intelligence
Computational linguistics
Computer graphics
Database management
Digital libraries
Human-computer interaction
Hypermedia
Hypertext
Image processing
Information storage and retrieval
Information highway/infrastructure, Internet
Networks, networking, networked information
See ei.cs.vt.edu/~mm/s97sspace/RelatedAreas_039.html
Send questions and comments to: Karen Lemone