Diapositive 1

1
The Comets Inspector Towards Run Time
Plasticity Control based on a Semantic Network
Alexandre Demeure, Gaëlle Calvary, Joelle Coutaz,
Jean Vanderdonckt Université de Grenoble,
France Université catholique de Louvain, Belgium
2
The problem

• UI plasticity adaptation of a UI
• to its context of use ltUser, Platform,
  Environmentgt
• while preserving usability
• E.g Choice of a month - Distinctive criteria

- Various
- Variable
- Unforeseeable
Task
Navigation
Genericity
3
Objectives

• Provide models, methods and tools for
• Describing and capitalizing UIs
• In an extensible way
• incorporating tailored UIs
• Supporting plasticity questions both at design
  and run time
• Which UIs are HTML versions equivalent to this
  UI?
• Is this UI decomposable into smaller ones?
• Is there a tailored UI for choosing a month?
  (pattern?)
• What is the common functionality supported by
  these 2 UIs?

4
State of the art

• Description

Choice in a known set Type a TYPE S_poss
set of Type S_eff set of Type min, max
Integer Constraints S_eff ? S_poss S_eff ?
min max User Task Specify S_eff
Domain Task
Languages XAML, XUL, UsiXML,
Abstract UI
Concrete UI
Final UI
5
State of the art

• Capitalization
• Taxonomy of interactors
• Rarely explicit
• Abstraction levels not explicit
• Not easily extensible
• Semantic network
• a graphic notation for representing knowledge in
  patterns of interconnected nodes and arcs.
  Computer implementations of semantic networks
  were first developed for artificial intelligence
  and machine translation, but earlier versions
  have long been used in philosophy, psychology,
  and linguistics (Sowa)
• Explicit the UI meta models
• Appropriate for extension
• Easy to represent abstraction levels

6
Approach

• Based on a semantic network for
• Capitalizing UIs
• Supporting plasticity both at design and run time
• Semantic network for plasticity
• Node ? UI models
• Edge(arcs) ? Relationships between UI models

7
Semantic network for plasticity

• Nodes
• Describe an interactive system at any level of
  abstraction
• Relationships
• Describe how a node is related to another one

Is deduced from
CT
AUI
CUI
FUI
Choice in a known set
Choice of months
8
Semantic network for plasticity

• Choice at CT level

Choice in a known set Type a TYPE S_poss
set of Type S_eff set of Type min, max
Integer Constraints S_eff ? S_poss S_eff ?
min max User Task Specify S_eff
9
Semantic network for plasticity

• Interleaving at AUI level

S_IS set of OBJECT
10
Plasticity questions
11
Plasticity questions
12
Plasticity questions
13
Plasticity questions
14
Plasticity questions
Is there a choice tailored for months?
15
Plasticity questions
Are there pie menu equivalent UIs without
navigation?
16
Plasticity questions
What are the possible TK choices?
17
Demonstrator

• Comet Inspector
• Comet Polymorphic interactors

18
Conclusion

• Contribution a semantic network that
• supports plasticity questions both at design and
  run-time
• is opened to tailored UIs
• Can be located inside and outside interactors
• Strength for plasticity
• Easy to extend Just add nodes and edges
• Enlarges plasticity without extra programmation
  The effort is set on the exploration mechanisms
• Preserves plasticity questions
• Drawbacks
• People must agree on a common vocabulary and
  structure
• Who controls the network?

19
Perspectives

• Theoretical
• Agreement on
• The UI levels of abstraction
• The UI relationships
• The structure of the semantic network
• Language to describe UIs and UI transformations
• Language to build plasticity queries
• How to include patterns
• Practical
• Semantic network server
• Visualization techniques for guiding the user
  (designer and/or final user)META-UI
• Integration to Comets (plastic interactors)
• Usability of the approach

20
Thanks

• QUESTIONS?

21
Semantic network for plasticity

• Who use?
• Conceptors
• design time Build different version
• Run-time Express substitution
• Users
• Personalize the interface (change this radio
  button choice with)
• Who extend?
• Conceptors new systems or new way to present
• Users End user programming (on the fly
  application building)

22
Semantic network for plasticity

• Nodes
• Describe an interactive system at any level of
  abstraction
• Example Choice at the CT level

Is deduced from
CT
AUI
CUI
FUI
Choice in a known set Type a TYPE S_poss
set of Type S_eff set of Type min, max
Integer Constraints S_eff ? S_poss S_eff ?
min max User Task Specify S_eff
23
Semantic network for plasticity

• Relationships
• Describe how an interactive system is related to
  an other.

24
Plasticity questions
25
Semantic network for plasticity

• Tailored UI
• Capitalizing the know-how

26
Semantic network for plasticity

• What are all implementation of a choice?
• Is there a choice tailored for months?
• What is the closest version of a TK pie-menu in
  HTML?
• What is the greatest common functionality between
  this 2 systems?
• What are equivalents to pie menu ensuring no
  navigation?
• Could this system be decomposed into smaller
  ones?
• What are the possible TK choices for choosing a
  month?