The Internet connects large numbers of people and large stores of information in decentralized, loosely organized structures, and we are seeing new forms of large-scale interaction emerge as a result. Where the interactive infrastructure workshop is focused on individuals working in a world rich with computing devices, this workshop focuses on interaction among large numbers of people who may or may not know each other. This workshop aims to investigate the technical research for augmenting human-human and human-computer interaction in such large scale, large population environments.
The way people interact in such spaces clearly offers new opportunities and challenges. New terms, such as “collective intelligence, “games with a purpose,” “social media,” and even “cloud computing,” point to new forms of interaction possible only in these environments. Indeed, finding new people to socialize with or finding the on-demand expertise necessary to solve an everyday problem are increasingly standard interactions. The first commercial pervasive or ubiquitous computing applications, such as small devices that provide peripheral awareness of others’ activities, point towards new kinds of large-scale interactions.
HCI as a field, with its commitment to human interaction as the locus of activity, needs to examine these new application spaces, with an eye to the enabling systems, mechanisms and middleware layers and base infrastructures necessary to enable these possibilities. It is also likely to be the area of Computer Science that will lead in the understanding of the significant issues and tradeoffs in next-generation environments. Indeed, HCI is the field most likely to understand that the social and collaborative nature of interaction is a prime requirement for these systems and also an outcome of those systems.
The purpose of the workshop is to enable interaction at ultra-large scale. Therefore, the topics largely include tools and infrastructures to foster interactions at Internet-scale and at pervasive-scale. We would also examine how to understand user activities and patterns at this scale as well.
These might include discussions of:
- Key components, services, and middleware layers that are critical for next generation user environments with an emphasis on scale of interaction. Services might include, for example, location awareness, sensor fusion, identification and identity services, activity recognition, user profiling, and incentive mechanisms. Middleware might include sets of services or components for collective intelligence, crowd-sourcing, and other social computing applications. These may, in turn, be based on infrastructures, upon which research has become dependent, such as language modeling, user modeling, data mining, information retrieval, and network visualizations.
- Research systems to create the next generation of facilities for such human-human interaction. Recent research has furthered the understanding of what kinds of interactions occur on the Net.
- Provisioning long tail content and knowledge sharing (e.g., for medical information, finding people, Q&A). As systems scale, it has become clear that people can find help, expertise, and even knowledge at a distance. Obtaining this help, expertise, and knowledge would have earlier required substantial effort, and the constant availability of enormous numbers of potential knowledge sources. Clearly new mechanisms for knowledge distillation, new ways to distribute and maintain expertise, and new ways to find and access knowledge are potential topics for further research.
- Creating new forms of incentive structures, especially those computationally based. Preliminary work has shown the utility of new forms of motivation and reward, including reputation systems, recommendation systems, systems that motivate through entertainment, and peer-based production. Creating facilities for large-scale experiments could also be a topic of discussion.
- Facilities to blend the real and the digital. As people increasingly become dependent on systems and as systems become pervasive, we will live in multiple social worlds, some will be digital and some will be partially digital and partially real-world. Facilities might include identification and authorization services.
- Usable privacy and security mechanisms must be part of any human-centered infrastructure. As collaborative information access continues and as the amount of digital traces for individuals grow, so do privacy and security risks for users.
- Considering what is appropriate evaluations for technical research within HCI, especially with regard to systems that may be difficult to deploy at scale.
It is important to note that all of these mechanisms, systems, middleware, and infrastructure layers are likely to be part of next-generation computational environments. Users will live in all of these simultaneously, and research must consider how these fit together in layered architectures that promote usability (in the narrow sense) and usability (as part of a socio-technical context).
Mark Ackerman – University of Michigan
Mark Newman – University of Michigan
Keith Edwards – Georgia Tech
Scott Klemmer – Stanford
Date: October 25-27, 2010
Location: Chicago, IL