This project will develop techniques that will allow users to help each other create and maintain configurations of complex, pervasive computing and communication environments. As personal computing environments become ever more complex, growing to include not just desktop and laptop computers, but also mobile phones, media devices, sensors, and more, configuration tasks grow in importance and difficulty. In particular, it becomes challenging for end-users to create, understand, and maintain the hardware and software configurations that allow them to carry out the activities that matter to them. Moreover, this problem will only get worse with time as computing environments grow to include the hundreds or even thousands of different devices and software services that have been forecast by computer scientists.
The approach taken in this project derives from the observation that, even as each user may have specific devices, services, and preferences that make it difficult for her to find information relevant to their particular needs, there frequently exists some other user, somewhere, who has experienced and solved a similar problem. This other user's knowledge would doubtless be of great benefit to the first user, but existing tools for seeking help and modifying configurations do not make it easy for such information exchange to take place. An important goal, then, is to match each user with the knowledge she needs in order to accomplish the configuration tasks facing her, on the assumption that such knowledge resides with some other user with a similar system.
This project will address a number of challenges, including: 1) How can "configuration knowledge" be identified and made available without placing undue burden on the individuals who possess it? 2) How can help-seekers be presented with information in a way that allows them to act on it with minimal effort and likelihood of error? 3) How can the complexity of large spaces of possible configurations be reduced to only the dimensions that matter for users' decision-making? In order to address these challenges, this project will develop the Collaborative Configuration Service (CCS) - a general service that collects configuration information from various users of a particular system and matches similar users with each other for the purpose of providing help. Through an iterative study-build-evaluate process, research will construct and refine CCS by adapting it successively to three different user communities: users of an ambient information device (Chumby), users of an open source software-based home media system (MythTV), and people representing the "early majority" of home media networking users.
Supporting users to gain control and receive help with the configuration and operation of open, evolving pervasive computing environments will lower the barrier to the adoption of those environments. This will lead to benefits to both the end-users themselves and to the companies for whom an open marketplace for pervasive computing services and components will present opportunities for competition and technological innovation.
