Multimodality and contextaware adaptation - PowerPoint PPT Presentation

1 / 48
About This Presentation
Title:

Multimodality and contextaware adaptation

Description:

Universit catholique de Louvain (UCL) Belgian Laboratory of Computer-Human Interaction ... Two examples of adaptation due to context-awareness. Adaptability ... – PowerPoint PPT presentation

Number of Views:65
Avg rating:3.0/5.0
Slides: 49
Provided by: jeanvand
Category:

less

Transcript and Presenter's Notes

Title: Multimodality and contextaware adaptation


1
Multimodality and context-aware adaptation
  • Quentin Limbourg, Jean Vanderdonckt

2
Context-awareness
  • What is in the context of use?
  • User
  • Platform
  • Environment
  • Organisation
  • Socio-psychological factors
  • Two examples of adaptation due to
    context-awareness
  • Adaptability adaptation by the user
  • Adaptivity adaptation by the system

3
Example of adaptability
  • Personalisation
  • Requires many editing
  • Is hard to achieve
  • Does not necessarily factor out common parts
  • Does not necessarily separate fixed parts from
    variable parts
  • Goal
  • To have a system for automated generation of UIs
  • With support of personalisation
  • That can generate multiple variants of the same
    UI
  • For this purpose, we use
  • A feature model
  • Generative Programming

4
Whats a Feature Model?
Exclusive
Optional
C
C
F1
F1
F2
Alternate
Dependencies
1.
C
2.
F1
F2
5
Example of a Feature Model
6
Feature model
Model Specifications
C
C
0
1
1
F1
F2
F3
F1
F2
F3
1
1
0
0
F2a
F2b
F3a
F3b
F2a
F2b
F3a
F3b
ltCgt ltF1gt0lt/F1gt ltF2gt1
ltF2agt1lt/F2agt ltF2bgt0lt/F2bgt lt/F2gt
ltF3gt1 ltF3agt0lt/F3agt ltF3bgt1lt/F3bgt
lt/F3gt lt/Cgt
USIXML Specifications
7
Generative Programming
8
Example of adaptivity
  • Computer-aided surgery by mixed reality

9
Multi-platform
  • Extending approaches for multiple contexts of use
  • Context is now restrictedto one computing
    platformat a time

10
Multi-platform
  • Traditional approach last level only

11
Multi-platform
  • Traditional approach Atomica

12
Multi-platform
UIML,1999
  • Shortcomings of the traditional approach (ctd)
  • Repetition of development and maintenance

Application 1
Application 2
Application 3
13
Multi-platform
Translation relationship Transformation at
the same level of abstraction
  • Physical level deployment

14
Multi-platform
  • Physical level deployment Galaxy

Platform 1 Personal Computer
Platform 2 Macintosh Mac OS
Platform 3 Workstation OSF/Motif
Physical User Interface
Physical User Interface
Physical User Interface
Windows UI code
Macintosh UI code
OSF/Motif UI code
15
Multi-platform
  • Physical level deployment Galaxy

16
Multi-platform
  • Logical level deployment

17
Multi-platform
  • Logical level deployment UIML

18
Multi-platformUIML Skeleton Part 1
Abrams et al., 1999
lt?xml version"1.0" ... ?gt ltuiml version"2.0"gt
lt/uimlgt
  • What parts comprise the UI whats their
    relationship?

19
Multi-platform
  • Task Concepts level deployment

20
Multi-platform
Thevenin,2001
  • Multiple levels of translation/reification

21
Multi-platform
  • Translation at Task Concept level

Platform 2
Task Concepts
Logical User Interface
Physical User Interface
Final User Interface
22
Multi-platform
  • Translation at Task Concept level
  • Task model primitives
  • Sub-tasks (nodes)
  • Remove sub-task for different reasons
  • Add sub-task unlikely to appear
  • Modify sub-task switch, reshuffle, relax
    constraints
  • Hierarchical temporal operators (edges)
  • Modify temporal operator reordering
  • Cutting sub-trees
  • Concepts model primitives

23
Multi-platform
  • Task translation

gtgt

24
Multi-platform
Märtin,1990
  • Concepts removal CT-UIMS

25
Multi-platform
  • Translations TC logical levels

Platform 2
Task Concepts
Logical User Interface
Physical User Interface
Final User Interface
26
Multi-platform
  • Translations TC logical levels

27
Multi-platform
  • Translations TC logical levels

28
Multi-platform
  • Translations TC logical levels

29
Multi-platform
  • Translations TC logical physical levels

Platform 2
Task Concepts
Logical User Interface
Physical User Interface
Final User Interface
30
Multi-platform
  • Translations TC logical physical levels
  • TC logical remain constant over time or
    coordinated with changes
  • Graceful degradation of presentation
  • Widgets replacement, reduction, resizing
  • Text text summarization, sentence elision,
    automatic summary
  • Images color reduction, cropping, resizing,
  • Animation reduction to still images,
  • Video reduction of frame, refreshing,

31
Multi-platform
  • Graceful degradation of presentation
  • Widgets

32
Multi-platform
  • Graceful degradation of presen-tation
  • Images

33
Multi-platform
  • Graceful degradation of presentation
  • Images

34
Multi-platform
  • Translations combined at all levels

Platform 2
Task Concepts
Logical User Interface
Physical User Interface
Final User Interface
35
Multi-platform
  • Multiple transformations possible

Platform 2
Final User Interface
36
Multi-platform
Calvary, Coutaz Thevenin, 2001
  • Plastic UI UI that can accomodate multiple
    changes of context (here, platforms) while
    preserving some usability properties

Platform 2
Platform 3
Final User Interface
Final User Interface
37
Multi-platform
Grolaux, 2001
  • Plastic UI Virtual keyboard

38
Multi-platform
Grolaux, Van Roy, Vanderdonckt, 2002
  • Plastic UI FlexClock

39
Multi-platform
  • Plastic UI adaptable bounded value

40
Multi-platform
Abstraction relationship Transformation into
amore abstract level ofabstraction
  • Reverse engineering

Platform 1
Platform 2
Task Concepts
WebRevenge
ReversiXML
Logical User Interface
Logical User Interface
ReversiXML
Physical User Interface
Physical User Interface
Final User Interface
Final User Interface
41
XIML2WML generator
42
Multi-platform
Crossing relationship Transformation
betweendifferent levels ofabstraction
  • Reverse engineering

Bouillon Vanderdonckt,2002
Platform 1
Platform 2
Retargeting
Logical User Interface
Physical User Interface
Final User Interface
Final User Interface
43
Multi-directional User Interfaces
  • A UI is represented by a suite of models that are
  • Analysable
  • Editable
  • Transformable
  • Transformational approach
  • Any UI is subject to any transformation at any
    time
  • From one language to another
  • From one computing platform to another
  • From reverse engineering to forward engineering

44
Multi-directional User Interfaces
  • Model engineering
  • Forward engineering (top-down)
  • Reverse engineering (bottom-up)
  • Bidirectional engineering (both ways)
  • Model transformations
  • Reification, abstraction
  • Translation
  • Crossing (retargeting, revamping,)

45
Multiple levels of consideration
  • Channels plus
  • User behaviour
  • User expectations
  • User experience
  • User groups
  • Physical constraints
  • Environment
  • Modalities plus
  • Market segments
  • Time of usage
  • Place of usage
  • Network, Wi-fi
  • Various tasks
  • Platforms plus
  • Modalities of interaction
  • Mono vs multimodality
  • Devices plus
  • Software characteristics
  • Hardware characteristics
  • Languages programming, markup
  • Tools renderer, interpreter, compiler

46
USIXML
47
An overview of USIXML tools
  • Running prototypes
  • GrafiXML CUI Hi-Fi editor Code Generator (Java
    Swing, XHTML)
  • FormiXML CUI Hi-Fi editor for Forms (XForms,
    HTML)
  • SketchiXML CUI Lo-Fi Sketching
  • VisiXML CUI Mid-Fi (MS Visio Plug-in)
  • FlashiXML CUI renderer for SVF
  • ReversiXML reverse engineering from HTML to CUI,
    AUI
  • TransformiXML API transformation tool
  • Teresa (F. Paternò, CUI level to multi FUI
    generation)

Demo
Demo
Demo
48
Thank you very much for your attention!
Credits
Write a Comment
User Comments (0)
About PowerShow.com