Catálogo de publicaciones - libros

Despite the availability of several task and personal information management tools, an appropriate support to human multitasking at work is still lacking. Supporting multitasking behavior entails capturing and modeling this behavior. In this paper, we refine an approach to model multitasking behavior in organizations, through an ontology based on two interrelated primitives; and The main contributions of the proposed ontology are: (1) enable the discovery of scheduling heuristics combining personal and inter-personal elements, (2) enable bottom-up discovery of tasks and (3) suggest a flexible system architecture for multitasking support. The first two contributions are illustrated through a case study.

Traditional task-elicitation techniques provide prepared structures for acquiring and representing knowledge about user tasks. As different users might perceive work tasks quite differently, normative elicitation and representation schemes do not necessarily lead to accurate support of individual users. If the individual perception of tasks should guide the development of user interfaces personal constructs have to be taken into account. They can be elicited through repertory grids: Personal work content and task-relevant information emerge in the course of structured interviews and can be transformed to conventional representation schemes, even for execution and prototyping. In this paper we introduce an elicitation procedure based on repertory grids and its embodiment in a working user-centered and task-based design approach.

In this paper, we describe the Cs Inspector, a software tool providing user interface designers and developers with a semantic network in order to control the plasticity of their User Interfaces (UI) at run-time. Thanks to a set of predefined relationships, the semantic network links together various concepts ranging from the final UI (i.e., the UI described in terms of techno lo gical spaces) to the concrete and abstract UIs (i.e., the UI respectively described in terms of concrete interaction objects independently of any technological space, and abstract individual components and containers independently of any interaction modality) up to the tasks and concepts of the interactive system. In this way, plasticity can be addressed at four levels of abstraction (task and concepts, abstract, concrete, and final user interface) for forward, reverse, and lateral engineering. The end user exploits the semantic network at run time to adapt his/her UI to another context of use by identifying, selecting, and applying plasticity suitable operations.

This paper describes a model-based development process for context-aware user interfaces. The development process consists of the specification and updates of several models followed by the generation and evaluation of a prototype. A generic runtime architecture will be presented supporting distinct prototyping renderers at different abstraction levels in order to support prototyped development during the whole design cycle of the context-aware user interface. To clarify the functioning of the architecture, a case study will be presented, demonstrating the possibilities of this prototype-driven development approach.