No abstract

Abstract.Many different task modeling methods exist. In this paper, we discuss 1) ingredients common to most task models, 2) how task modeling relates to the design of user interfaces, and 3) our proposed ontology for task analysis. We then show our task analysis tool that is based on the ontology. It is our belief that task models should be based on an ontology that describes the relevant concepts and the relationships between them, independently of any used graphical representations. Such an ontology helps to understand the different task modeling methods and it can also be operationalized for use in tools.

No abstract

Groupware Task Analysis is a task analysis method that deals with the context of use of a system in the broadest sense. The task world is seen from three viewpoints that deal with different aspects of the world. The processes of GTA and their background are described in detail. In addition a task analysis tool EUTERPE is described. EUTERPE is based on GTA and allows capturing of the task models and provides some basic analysis primitives. INTRODUCTIONAnalyzing a complex system means analyzing the world in which the system functions, or the "context of use", which comprises (according to standards like [1]) -the users; -the tasks; -the equipment (hardware, software, and materials); -the social environment; -the physical environment. If we want to design systems for the context of use, we need to take these different aspects of the task world into consideration. In traditional literature on task analysis from the HCI (Human-Computer Interaction) mainstream, the focus is mostly on users, tasks, and software. Design approaches for GroupWare and CSCW (Computer Supported Collaborative Work), on the other hand, often focus on analyzing the world first of all from the point of view of the (physical and social) environment. In both cases, more recent developments at least include some aspects that belong to the other categories, but it still looks like one has to choose for either the one view or the other. Section 2 presents an idea of task analysis approaches from the classical HCI tradition, and, at the same time, provides our view on phases in task analysis. In Section 3 an ethnographic viewpoint, as frequently applied to the design of CSCW systems, is presented, where phases in the analysis process are hardly considered. As the result of combining approaches from both HCI and CSCW design, we developed our GTA (Groupware Task Analysis) framework of modeling task knowledge, which we will describe in section 4. Brigitte Jordan [2], though originally working from an ethnographic approach and focusing on GroupWare applications, provides a view on analyzing knowledge of the task world that is broad enough to cover most of the context of use as now defined by the above mentioned ISO document. We will illustrate Jordan's view in Section 5, distinguishing two factors: -sources of knowledge: (1) individual knowledge, and (2) group; -levels of communicability: (a) explicit, and (b) implicit. Based on applying Jordan's 2 * 2 framework in actual design processes for large industrial and government interactive systems, and expanding the two factors from dichotomies to continuous dimensions, we describe a twodimensional framework to analyze the different relevant sources of knowledge of the context of use. This framework provides a map of knowledge sources, that assists us to identify the different techniques that we might need in order to collect information and structure this into a model of the task world. Section 6 describe EUTERPE, a task analysis tool based on an ontology derived from GTA. The model used by the tool ...

Abstract. Usability has gained a lot of attention in the design community and it is one of main goals of every design project. Evaluating usability is usually done with end-users after a prototype has been built and there are not many techniques available that allow usability evaluation during the early design phases. Current dialog modeling techniques generally do not deal with usability aspects, as they are often functional based models, dealing only with states and state changes. This paper investigates how usability aspects can be incorporated into dialog models so that usability can be evaluated during the design process without doing usage tests. A set of measurable properties is given which together could give an indication about the usability of the design, This way, some usability aspects can be covered early in the design process without the need for an executable prototype or end-users.