This paper presents a formal model to check the interaction plasticity on a user interface (UI). An interaction is seen as an implementation (achievement) of a user task by means of interaction devices and modes of a given platform. The interaction plasticity is the ability of UI to support several interactions to perform the same task. In this work, two task models, containing different sets of interactions, are observed to check if they describe interactions that perform the same task. Each task model is represented by a labelled state-transitions system (lts). Due to the use of different interaction modes and devices, the obtained lts have different set of labels. Weak bisimulation relationship is revisited to handle these transition systems by defining a relation on labels. This relation is borrowed from an ontology of interaction modes and devices. Model checking techniques are set up to automatically establish such a bi-simulation. A case study is used to illustrate how the approach works.

The considerable and signiÞcant progress achieved in the design and development of new interaction devices between man and machine has enabled the emergence of various powerful and efficient input and/or output devices. Each of these new devices brings speciÞc interaction modes. With the emergence of these devices, new interaction techniques and modes arise and new interaction capabilities are offered. New user interfaces need to be designed or former ones need to evolve. The design of so called plastic user interfaces contributes to handling such evolutions. The key requirement for the design of such a user interface is that the new obtained user interface shall be adapted to the application and have, at least, the same behavior as the previous (adapted) one. This paper proposes to address the problem of user interface evolution due to the introduction of new interaction devices and/or new interaction modes. More, precisely, we are interested by the study of the design process of a user interface resulting from the evolution of a former user interface due to the introduction of new devices and/or new interaction capabilities. We consider that interface behaviors are described by labelled transition systems and comparison between user interfaces is handled by an extended deÞnition of the bi-simulation relationship to compare user interface behaviors when interaction modes are replaced by new ones.

No abstract