Other terms for adaptive graphics:

• Personalised Graphics
• Customised Graphics
• Data Driven Graphics

Classification of E-Publishing

• E-Publishing
  • Static
  • Dynamic
    • Offline -- initiated by publisher
    • Online -- initiated by reader

Plot of cost vs. actuality, saturation curve: static, offline, online

Multi-stage publishing pipeline, several intermediate formats before final display

Layered Architecture

• Application - SVG,XHTML
• Presentation - CSS,FO
• Transformation -- XSLT, DSSSL, fxt
• Content

Content and Presentation Separation.

For documents this splits text from font/color/text properties

Raster images have both content and presentation.

Vector images, more amenable to changes: filter, sizing, deformation, removing the background, etc. The latter example is more compelling, others can be done with standard graphic manipulations.

• Content - abstract description of image
• Transformation -- select add, edit, delete parts, transforms into user format
• Presentation - color schemes

Adaptation Properties

• Quantity changes -- date/time, e.g. time-zone, e.g. when user made the request
• Quality changes -- based on user preferences, e.g. language, location

• Static layer -- traditionally the background, but not necessarily; doesn't change
• Dynamic layer -- contains a number of overlays; can change

Examples

Example of quality changes: online comic being translated into other languages.

Example of quantity changes: entertainment maps, showing lectures, movies, exhibitions, etc.

Tokyo subway: static layer is lines, dynamic layer is time until next train indicators.

Can combine quality and quantity changes together, e.g. weather maps. Different weather on different days; different map types (geo vs. political); translation

BioML, Genetics. Browser for genetic information. Representational tree view, highlights gene sequence. Can be improved by generating SVG view of interaction between amino acid sequences.

Medical education. English/Latin labels on images, overlay of X-rays

MusicML. Dynamically generating musical notation from the XML description. Can also generate sound.

Audio

• Content -- transcript (interview), music notation
• Transformation -- add, edit, delete tracks and parts
• Presentation -- voice, instrument?, speed?

No ideas for video as yet.

Summary

• Separation (content/presentation) is difficult, but potentially as useful as for text
• Definition of level of abstraction
• Classification of dynamic and static parts of the image

Questions

Some discussion about adaptive video: screen resolutions, blue-screening, DVD options, etc.