Modelling multi-user tasks

Task models are the meeting point about multiple views of an interactive software application because their full development often require the involvement of many expertises: the designer, the expert of the application domain, the end user, the software developer and so on. The need for this possible multiple contributions is derived also for their possible multiple use: o To improve understanding of the application domain, in this case designers develop task models corresponding to the existing application and the modelling work can stimulate the request of many clarifications whose need may not be perceived just reading textual descriptions or documents; o To support effective design, in this case designers develop an envisioned task model of a new application incorporating requirements received (for example to use some specific technology or to overcome some limitations for end user that occur in current applications) and the various information contained in the new task model can support the design and implementation of a concrete user interface; o To perform usability evaluation, to compare how users interact with an application and the related task model (Lecerof & Paternò 1999) can be useful to understand limitations of the current design, whose improvement may imply changes in the corresponding task model. Despite of the increasing recognition of the importance of task models in the HCI area there is a lack of engineered tools supporting their use. This limitation is a consequence of the lack of structured methods and precise notations for task models. In this paper we introduce and discuss how we are solving such a problem by developing an automatic environment for task models specified in the ConcurTaskTrees notation (Paternò 1999).