7.1 Using the RM

The current model will be helpful to analyze and compare existing systems, and it certainly will provide valuable guidance for the development of new IMMPSs. For illustration, we derive from the RM a list of questions which one may answer to set up a profile for a concrete system S.

Input:

• What are the presentation goals and commands to be handled by S?
• What kinds of application data/knowledge are assumed?
• Are there any further input parameters for S?

Output:

• What kinds of presentations are delivered?
  • Which types of media objects, and which combinations?
  • Which modalities?
• Is the realization of media objects uncoupled from displaying them?

Generation:

• Which of the subtasks specified by the layers (and their components) are handled by S? In case if:
  • Which mechanisms are used to handle them?
  • Does S assume a fixed order in which subtasks are carried out, or is the ordering dynamically determined based on some knowledge?
  • How well are these tasks performed?
• Which knowledge sources are exploited?
  • Which sources are private, which are shared?

...

Performance:

• To what extent are variations in the input to S reflected as variations in the output?
• How much interaction between system components is required when generating e.g., cross-media references?

...