Information Technology Reference
In-Depth Information
For the individual, UI adaptation to the interaction device is performed
automatically: photo taking is only proposed to the person if his cell phone is
equipped with a camera, see Figure 11.4b. The localization of the incident is
automatic when the cell phone has a GPS, see Figure 11.4c. Otherwise it is
manually done by checking on a map whether the cell phone has a map of the place,
see Figure 11.4d or, if not, by specifying the address of the place, see Figure 11.4e.
If after connecting his PC to Sedan-Bouillon, Lionel connects to this same site
from a PDA, then this double connection is detected and a redistribution proposal is
made to him: a
negotiation
UI pops up that explains to Lionel that he can distribute
the site, as he sees fit, between the PC and the PDA, see Figure 11.5b. The
redistribution is done according to workspaces (the zones). By ticking the
appropriate check boxes, Lionel specifies the allocation of the various workspaces
to the different platforms. This additional UI enables the dynamicity to be placed
under human control. These supervision UIs, which enable users to program their
ambient spaces, are called “meta-UIs” by [COU 06]. The redistribution then
proceeds, recomposing the UI under the control of the user, see Figure 11.5c.
In addition to dynamic adaptation triggered by the dynamic discovery of
interaction resources, two other variables need to be considered:
- the variability of
information
provided by services that appear and
disappear opportunistically; and
- the variability of the user's
intention
as the result of a change of place or the
arrival of a particular information.
11.2.3.
User control
It can be important for the user to have control of the UI adaptation process. One
possible approach is so-called end-user programming, which makes it possible for
“non-ICT specialists” to develop their own services and systems. Although end-user
programming was initiated 30 years ago [SMI 77], this area of research is being
rediscovered in the context of ambient intelligence.
Research in end-user programming first sought to define simplified notations:
scripting textual (
HyperTalk
) or graphical languages (
Visual AgenTalk
) [REP 04],
programming by demonstration and by example [CYP 93], as well as the
construction of macros (similar to emacs). These techniques were essentially applied
to targeted areas such as Computer-Aided Design [GIR 92], spreadsheets [BUR 03],
bio-informatics [LET 05] or, like Alice [CON 97] and HANDS [PAN 02], to the
learning of programming. Recently, research has opened up to the Web, with the
possibility for users to build macros: with Koala, a user saves his actions on a
