K-MADe Story

Current References

For the tool: M. Baron, V. Lucquiaud, D. Autard, et D. L. Scapin (2006). K-MADe : un environnement pour le noyau du modèle de description de l’activité. submitted to IHM 2006. Proceedings of the 18th French-speaking conference on Human-computer interaction (Conférence Francophone sur l'Interaction Homme- Machine) ; Montréal ; Canada.

For the model : M. Lucquiaud, V. (2005). Proposition d’un noyau et d’une structure pour les modèles de tâches orientés utilisateurs ; Proceedings of the 17th French-speaking conference on Human-computer interaction (Conférence Francophone sur l'Interaction HommeMachine) ; Toulouse ; France ; pp 83-90 ; 2005.

History and past References

The initial work concerning task modelling for user interfaces (Scapin, 1988) had several objectives: to consider the way in which users represents their task, not the logic of the data-processing, or the “prescribed” task; to take into account the conceptual and semantic aspects, not only syntactic and lexical aspects; to obtain a structuring of the tasks in a uniform way; to carry out descriptions from a declarative point of view (state of things) as well as procedural (way of reaching these states); to allow parallelism and not only sequential representation(synchronization of the tasks); and to be computable.

From these first ideas and various requirements, a first task analysis model (“MAD”: Méthode Analytique de Description) was proposed (Scapin, D.L. & Pierret-Golbreich, C., 1989, 1990) at the intersection of ergonomics, computer science, and artificial intelligence. Moreover, the tool EMAD was developed (Pierret, C., Delouis, I. & Scapin, D.L., 1989); Delouis I., Pierret C., 1991).

In parallel, a method for task information gathering was defined (Sebillotte S., 1991); it consists mainly of semi-directed interviews based on the “Why?” “How?” technique. Also, one practical methodology of task analysis for extracting the relevant characteristics useful for the design of interfaces was also proposed (Sebillotte, S. & Scapin, D.L., 1994).

Since then the work has been continued according to an iterative process: development; for example: implementation of the model in an object oriented fashion, organization of the tasks at various levels of abstraction, validation in the field (house automation, management of ship fires, air traffic control, railway regulation, etc), modifications, validations again (further references are available upon request).

Research work (Gamboa-Rodriguez, F. & Scapin, D.L., 1997) also focused on the design and specification of interfaces. Two models were designed: a task model (MAD *) and a model for conceptual interface design (ICS), as well as procedures for moving from one model to the other, and links with the low level interface aspects (via the toolbox ILOG-Views). The work led to the implementation of the task analysis editor EMAD * and to a software design environment (ALACIE).