|
Home | First International Workshop for Computer Support for Human Tasks and Activities
INTRODUCTION
Unfortunately, existing computer systems provide inadequate support for human tasks and activities. By offering only single applications and files as first class entities, current systems are oblivious to a notion of a user task spanning multiple applications, services, and information sources. For example, ordinary office work such as writing a memo may require bringing up a text editor, a spreadsheet, a web browser, and an email client, each viewing the documents, web pages, and emails that are relevant for the task. When shifting to a different task, the user must save his work on the previous task and bring up the relevant applications and information sources for the new task. However, human tasks are characterized by spontaneous change, interruptions, parallelism, and collaboration. Therefore, by lacking a first class notion of user task, current systems fail to support easy shifting between tasks.
This situation is worse in work environments where users, rather than sitting at a desk all the time, would like to use whatever devices are available at each location for carrying out their work and for impromptu collaboration. Again, current systems provide inadequate support for such a requirement. For instance, for the medical staff at a hospital, it is impractical to shift their attention from the focus of their task – a patient – in order to locate a device, login, and locate the relevant application/file to show to a colleague.
Additionally, there is currently little support for the overhead associated with managing tasks. For instance, busy users who receive dozens or hundreds of emails a day would like to have assistance relating those emails to each of their tasks, recognizing which are important, and which are urgent. Sophisticated assistance could include reacting to the contents of the emails, for instance, recognizing that a meeting is being requested by a colleague and automatically checking the user’s calendar for availability.
THE WORKSHOP
This workshop aims at exploring issues related to improving the computational support for human tasks and activities, with a special focus on infrastructures, software architectures, models of user tasks, and on the challenges associated with designing and implementing those.
The goals of this workshop are the following:
We seek contributions within, but not exclusively, the following subtopics of activity-based computing:
The Program Committee will appraise the quality of the submissions and between 10 and 15 participants will be admitted to the workshop.
PROGRAM COMMITTEE
RUNNING THE WORKSHOP
The format of the workshop will be for participants to briefly present their vision for future directions in this research area. This will take the first half day. In the afternoon we break into working groups based on the themes identified on beforehand, moderated by the workshop organizers. In a plenary session following the groups’ work each group will present their ideas and discuss them with all the workshop participants. In a brief concluding plenary session, all the participants will have the opportunity to provide feedback on the workshop and decide for any continuation at another venue.
ABOUT THE ORGANIZERS
Henrik B. Christensen’s research interests include software architecture, software engineering, pervasive computing, and teaching. As researcher at Centre for Pervasive Healthcare, he has been architect on the ABC framework that experiments with computer support for human activities and collaboration. His research includes conceptual frameworks, architectures, and programmers API’s for activity based computing, proactive discovery of activities, and techniques and tools for engineering of reliable pervasive services.
David Garlan's research interests include software architecture, ubiquitous computing, self-adaptive systems, formal methods, and software development environments. He is a principal investigator on the Project Aura, a large-scale research effort to explore the next generation of system infrastructure in support of pervasive, mobile computing, integrating research in networks, operating systems, software architecture, and user interfaces. He is also a participant in the Radar Project, which is developing highlevel cognitive assistance for computer users, including learning of user preferences and policies, and deep knowledge of specific task domains.
João Pedro Sousa's research interests include ubiquitous computing, self-adaptive systems, and software architecture. He is a researcher at CMU's Project Aura, focusing on the automatic configuration and reconfiguration of computational environments to support human tasks. His research includes modeling user needs and preferences; an architectural framework that supports automatic (re)configuration by enforcing task-specific adaptation policies; and measuring the user's cognitive overhead when using such systems.
REFERENCES
|