Information Technology Reference
In-Depth Information
within the Viatic.Mobilité project by sociologists and anthropologists who
considered it to be relatively simple to understand and handle, all the while being
relevant to the fact that, in principle, the user only rarely goes from one state to a
completely different one.
9.4.5.
Summary and discussion
Via these examples, we have shown that our approach could potentially enable
us to take into account the travel time experience of the user on several levels in
applications:
- as an element enabling the content of all the business processes associated with
the application to be distributed and defined;
- as a selection criterion, directly or via a business rule, enabling choices within
a business process to be defined and therefore within the conceptual models of the
application.
These examples also enable the current limitations of the approach to be
identified:
- The importance of the preliminary analysis phase, to distribute and adapt the
different business processes, which is based on a pluridisciplinary approach that is
often difficult to set up.
- The difficulty of exactly knowing the real category of the travel time
experience of the user; the category changes only being made via interactions
carried out by users, the other elements being ignored. Thus, talking to a neighbor
after having worked on a dossier is a typical example of a category change that is
ignored by the application as it is impossible to identify in the current state of our
works. Nonetheless, we have a certain number of preliminary ideas for the use of
existing sensors to take into account category changes, such as the use of sound
sensors to identify passage to the
Peacock
category. All of these avenues still need
to be validated, however, as the choice of relevant sensors and the definition of
ensembles of associated values for each category represent an important focus of
research that we have only just started to address.
- Finally, there is a real issue with the characterization of categories and the
number of them. In the context of our works, we have used four categories of user
classification, but other works have put forward three [FIC 70] and nothing says that
there should not be more than four categories. With the characterization of
categories, the aim of which is to specifically define the criteria enabling a user to be
identified as being part such-and-such a category, the difficulty resides in the
definition of criteria that are discriminatory enough to avoid all ambiguity when a
category is associated with a user. The characterization and definition of the number
