Panel Session
Towards a Better Understanding of Context and
七月英文缩写Context-Awareness
Moderators
Gregory D. Abowd and Anind K. Dey
Graphics, Visualization and Usability Center and College of Computing,
Georgia Tech, Atlanta, GA, USA 30332-0280
上海暑假夏令营{abowd, anind}@cc.gatech.edu
Panelists
Peter J. Brown1, Nigel Davies2, Mark Smith3, Pete Steggles4
1 Computing Lab., The University of Kent at Canterbury, Canterbury, Kent, CT
2 7NF, UK
P.J.Brown@ukc.ac.uk
2 Computing Dept., Engineering Building, Lancaster University, Lancaster, LA1 4YR, UK
nigel@comp.lancs.ac.uk
3 Hewlett Packard Laboratories, Palo Alto, CA, USA
msmith@hplmts.
4 AT&T Laboratories Cambridge, 24a Trumpington Street, Cambridge CB2 1QA, UK
anzac day1 Introduction
When humans talk with humans, they are able to u implicit situational information, or context, to increa the conversational bandwidth. Unfortunately, this ability to convey ideas does not transfer well to humans interacting with computers. In tradi-tional interactive computing, urs have an impov
erished mechanism for providing input to computers. By improving the computer s access to context, we increa the richness of communication in human-computer interaction and make it possible to produce more uful computational rvices. The u of context is increasingly im-portant in the fields of handheld and ubiquitous computing, where the ur s context is changing rapidly. In this panel, we want to discuss some of the rearch challenges in understanding context and in developing context-aware applications.
We define context as any information that can be ud to characterize the situation of an entity, where an entity can be a person, place, or physical or computational ob-ject. We define context-awareness or context-aware computing as the u of context to provide task-relevant information and/or rvices to a ur. Three important context-awareness behaviours are the prentation of information and rvices to a ur, auto-
H.-W. Gellern(Ed.): HUC’99, LNCS 1707, pp. 304-307, 1999.
Springer-Verlag Berlin Heidelberg 1999
matic execution of a rvice, and tagging of context to information for later retrieval.Some of the main challenges in the area of context-aware computing are:
• the development of a taxonomy and uniform reprentation of context types;
• infrastructure to promote the design, implementation and evolution of context-
aware applications; and
• a discovery of compelling context-aware applications that assist our everyday inter-
actions with ubiquitous computational rvices.sat答案
2 Position Statements
Following are the position statements from each of the panelists.
会计证怎么考
2.1 Peter Brown (www.cs.ukc.ac.uk/rearch/infosys/mobicomp/Fieldwork/)My view is that future rearch must focus more on the whole solution rather than on part solutions. There are two strands to this. The first is general to most rearch.When I was working with a company to create a hypertext product, I was told "People do not want hypertext: they want solutions. Hypertext is only likely to be PART of an solution". I often quote this maxim, as one can substitute almost any rearch area for hypertext. Certainly it applies to context-awareness. The cond strand concerns
bringing together technologies. You could get any of the following opinions from individual rearchers in context-awareness:- the key is the underlying databa.- context-awareness is a type of information retrieval.- what we need to do is to solve the HCI issues.- we need a solid theoretical foundation.- if you want this to be real, you need to solve hard AI problems.- it is a systems/software engineering problem.- the area is waiting for the right hardware at the right price.- context-awareness is a specialized area of distributed systems.All of the are at least partly true. The real key, however, is bringing the technologies together.
My theme, therefore, is that currently we rearchers think too much in terms of our own pet technology. The reality is that this will only be part of context-awareness,and, if we are hoping for our rearch to be applied, this in turn will only be part of the solution. Certainly such success as we have had at the University of Kent at Canter-bury has been the result of people with different outlooks getting together, but we, like others, still have a long way to go.
305
Towards a Better Understanding of Context and Context-Awareness
306G.D. Abowd et al.
2.2 Nigel Davies ( www.guide.lancs.ac.uk)
Context nsitive systems and applications are tho that respond to changes in their environment. Typically the respons are designed to improve a systems' perform-ance or to make its behaviour more relevant to the situation in which it is being ud.
F or the purpos of this panel I propo to argue that in the near future all mobile systems will be context-nsitive: after all, why would you want a system which was not context nsitive? I will provide examples to justify this claim bad on the results of our work on the Lancaster GUIDE project. This project has developed and de-ployed a context-nsitive tour guide in the city of Lancaster which combines mobile computing hardware, context-nsitive applications and wireless communications to provide visitors to the city with context nsitive, interactive applications.混淆怎么读
A relevant publication is: Davies, N., K. Cheverst, K. Mitchell, and A. F riday. "Caches in the Air: Disminating Information in the Guide System" Proc. 2nd IEEE Workshop on Mobile Computing Systems and Applications (WMCSA '99), New Orleans, U.S., IEEE Press, Pages 11-19.
2.3 Mark Smith
英文铃声
Context aware computing has the potential to allow applications in emerging rvice bad computing architectures to provide completely new functionality. Client de-vices can u a wide variety of nsors to provide awareness of who the ur is, and the location and environment in which the application is being ud. Environment awareness can include such things as whether or not a ur is still holding a device, if the ur alone or with others, stationary or in motion, hot or cold, light or dark and so forth. Using this information, appliances and applications can be optimized and per-sonalized in ways that provide benefit to both technology providers and urs. To realize this vision, I e six broad rearch areas in context aware computing. They cut across many of the technical areas in rvers, rvices, transport and clients, both fixed and mobile, that make up future rvice centric system architectures.
1.Algorithms to render physical significance from nsor data. This is most obvi-
ous in applications such as biometric bad ur authentication, but it can become very challenging in areas of environmental awareness. For example, an algorithm that exploits nsor data to robustly determine something emingly simple such as if a device is still in the posssion of a ur can be difficult.
2.Technology for integrated nsors. Some nsors such as active pixel arrays can
currently be integrated onto CMOS chips without using special processing, but what about other nsors such as motion or humidity. How can nsors be de-signed to allow multiple functionality, for example an imaging nsor that can be ud for both biometrics and communication. Integrated optics is another chal-lenge both for capture and display.
3.Privacy issues. The go way beyond the curity and encryption issues we have
now. I would not be surprid if in the future entire conferences are given to this topic.
女生英文名大全带翻译4. Agents that make u of nsor data. How are agents to be constructed that sup-port dynamic persona in an application? How is the data that the agents u to be distributed? The are problems at the rver and transport level.武汉辅导
托福出国5. Personalization of applications and appliances. One of the greatest benefits of using context aware computing is the ability to u nsor data to personalize ap-plications and devices to conform to what the ur expects. This is from the con-sumer s point of view the killer app . The are problems at the application and client device level.
6. Capitalizing on the new spaces that context awareness enables. What customers will pay for, and
what not. What partnering will be necessary across the emerging rvice centric computing community that is forming in order to make context aware computing real.
2.4 Pete Steggles (/~pjs)
What is needed in context-aware computing is a sufficient kit of composable small rvices which really benefit from context-awareness. But with the right parts you can put together what I would call killer applications. We have just put together a system which combines follow-me desktops, camera tracking, text-to-speech, follow-me sound playing and 3D visualization to provide a recorded tour of our lab where we take a visitor round with a slide show following us, demo the technologies referred to in the show at the time, track the visitor with cameras and record the whole thing,giving the visitor a souvenir CD with a multimedia version of their tour with synchro-nized video. F or us, when we have fully deployed our system, that will be a killer app!Any description of the world that can be relevant to an application counts as a context type . In our rearch we have attempted to create a very detailed model of the environment to address this. Inevitably, with so few real context-aware applica-tions, there is doubt as to which facts will be uful and which won't. With respect to location, we have got a long way by attempting to model vague location-related facts using geometric containment and overlapping relationships. Really accurate location is 99% of the battle.307
Towards a Better Understanding of Context and Context-Awareness
