Context-Awareness and Context-Transparency as ... - Semantic Scholar

3 downloads 11956 Views 30KB Size Report
major HCI conferences and it indeed seems to be straight- forward that ... A context-aware cellular phone, for ... call and when it is appropriate to use a phone.
Context-Awareness and Context-Transparency as Orthogonal Concepts in HCI Christopher Lueg Department of Information Technology University of Zurich Winterthurerstrasse 190, CH-8057 Zurich, Switzerland [email protected] Abstract There is no free ride in enhancing context-awareness of artifacts as artifacts lack the ability to (re-) negotiate the meaning of particular situations. We therefore argue that research in context-awareness should be complemented by orthogonal research in what we call context-transparency.

1

Introduction

Context-aware applications have become a hot topic at major HCI conferences and it indeed seems to be straightforward that user-friendly applications should be able to adapt their functionality to the respective context in which they are being used. A context-aware cellular phone, for example, could use location data and the user’s current schedule to determine whether it is more appropriate to ring or to buzz (or even to vibrate) in order to notify the user of incoming calls. The phone even might decide to suppress notifications of less important calls and foster notifications of the important ones. In what follows we argue that there is no free ride in enhancing context-awareness as artifacts are unable to (re-) negotiate meaning. In order to complement research in context-awareness we propose to investigate what we call “context-transparency” and we illustrate some key aspects of context-transparency.

2

Degrees of context-awareness

In HCI research, the term context-awareness refers to the idea that artifacts should be able to determine in what context they are being used in order to adjust their functionality to the specific characteristics of that context. Perspectives in the ongoing context-awareness debate range from rather tool-oriented perspectives featuring the idea that it is within the user’s responsibility to use an artifact in an appropriate context to strong AI-related contextmodeling perspectives where it is up to the artifacts to recognize the appropriateness of a usage context. A rather simple example for the extreme “soft” position is the use of a hammer. According to rumours, to a hammer the world looks like consisting of nails, and it is commonsensical that it is up to the user to provide an appropriate “nail context” when using a hammer. Also,

when using regular cellular phones, users are able to use the phones only when they would like to make a phone call and when it is appropriate to use a phone. Examples for rather “strong” approaches in contextawareness are, apart from the aforementioned concept of a context-aware cellular phone, cooperative buildings that recognize whether a meeting or an informal get-together is taking place, electronic couches that provide occupants with context-specific information, electronic treshholds that surveille people entering a room in order to improve scheduling, and electronic beds taking care of the users’ needs when turning in bed.

3 Context and the negotiation of meaning Investigating context-awareness necessarily involves coping with the notion of context. However, rather than harping on the impracticability of the complete modeling of situations (e.g., [2]), we would like to draw the attention to some other aspects. When introducing the hammer example, we mentioned that it is up to the user to provide an appropriate “nail context” when using such a tool. This means that —in this particular case— there is no need for the artifact to be context-aware in order to adapt its functionality as the user already provides a context that fits the artifact’s functionality. The capability to provide appropriate context, however, is only one part of the story. The interesting thing with the term appropriate is that humans are not only capable of generating context in order to use an artifact as intended by the designer but humans are also able to use artifacts in other contexts than originally envisioned. This means that humans are capable of generating new —but no less appropriate— contexts. For example, a hammer can not only be used to drive nails but, among other things, a hammer can be used as a paper weight. Another example is a powerful computer on one’s desktop that can be a status symbol (e.g., [6]). From a rather humane perspective, the question is not so much how artifacts could be made more context-aware but how artifacts could be designed in such a way that the user is supported in fitting the artifact to his or her current situation. Referring to “situation” rather than to “context” allows to distinguish between aspects that are subject to continuous negotiation and aspects of the envi-

ronment that are rather persistent, i.e., aspects that can be sensed or measured so that it makes sense to incorporate them into a context model. The aforementioned example of a cooperative building is well-suited to illustrate the difference between “context” and “situation”. Within the building, a room’s electronic schedule, the number of persons in the room, and perhaps even the prevailing clothing are among the things that could be part of a “business context” model. However, even if a meeting perfectly fits such a model, the meaning of the meeting and thus the situation can still be informal precisely because the meaning is not static but under continuous negotiation among the participants; what started as business meeting can be turned into an informal get-together and vice versa. Another nice example for the negotiation of a situation can be found in the movie “Total Recall”. Upon arrival on Mars, the hero (Arnold Schwarzenegger) is tracked by evil guys. With a bit of luck, he makes it into a fully automated cab. The robotic driver recognizes that a new guest has entered his cab and —as expected in a “cab context”— the robot starts asking for the destination. Failing to recognize the dramatic escape situation, the robot wastes valuable time querying repeatedly for a destination (until kicked out of the cab). In terms of negotiation [6], the robot was unable to re-negotiate the meaning of the situation.

4

A quest for context-transparency

Artifact design can hardly account for continuously negotiated meaning. However, design can account for the human capability to negotiate the meaning of artifacts in a situation and to generate context ”on the fly”. Artifacts that are supportive in this respect qualify as what we call “context-transparent”. A cellular phone application (see figure 1) developed by Swiss Radio Pilatus is well-suited to illustrate some key aspects of context-transparency. The application is situated in the broadcast domain. Often, users would like to know what song is currently on air as he or she is interested in buying the CD. Radio Pilatus addressed the issue as follows: when a user would like to know the details of the song currently on air, he or she can request the details by sending a specific SMS (a short textual message that can be sent and received using cellular phones; such phones are almost omnipresent in Switzerland) to the radio station. The radio station immediately returns the details of the song. Furthermore, the user has the chance to order the CD directly by sending some other kind of SMS (given that the user has registered before and has provided a bank account or credit card number). The example nicely illustrates the following aspects of context-transparency. The appropriate usage context is created by the user being in a particular situation by assigning meaning to one aspect of the situation which is the music currently on air. In addition, the user assigns specific meaning to an artifact that is typically used for other purposes but the new meaning is in line with the

Figure 1: Details of the song currently on air requested via SMS (short textual messages that can be sent and received using cellular phones). Instant CD ordering via SMS is possible as well. Courtesy of Radio Pilatus, Switzerland. given functionality of the artifact as well. Future research in context-transparency is to investigate what aspects support and what aspects inhibit the creation of context in a particular situation, and how these insights could be incorporated into artifact design.

5 Related work Studies of the task-artifact cycle [1] illustrate that using artifacts influences in turn the users of these artifacts. Participatory design approaches (e.g., [4]) investigate how design can account for the side-effects of artifacts being introduced to work situations. Persuasive computing [3] explores how artifacts could be used to change a person’s attitudes or behaviors.

References [1] J. M. Carroll, W. A. Kellogg, and M. B. Rosson. The task-artifact cycle. In J. M. Carroll, editor, Designing Interaction. Psychology at the Human-Computer Interface, pages 74–102. Cambridge Univ. Press, 1991. [2] B. Edmonds. The pragmatic roots of context. In Proceedings of the Second International Conference on Modeling and Using Context, pages 119–132, 1999. [3] B. J. Fogg. Persuasive computers: perspectives and research directions. In Proceedings of the Annual ACM SIGCHI Conference on Human Factors in Computing Systems, pages 225–232. ACM Press, 1998. [4] J. Greenbaum and M. Kyng, editors. Design at Work: Cooperative Design of Computer Systems. Lawrence Erlbaum Associates, Hillsdale, New Jersey, 1991. [5] R. P. Online. Senderecherche via SMS - CD’s direkt bestellen. http://www.radio-pilatus.ch/sound/ smsrecherche.asp (last page visit 24/5/2000). [6] E. Wenger. Communities of Practice: Learning, Meaning, and Identity. Cambridge University Press, Cambridge, UK, 1998. First Paperback Edition 1999.

Suggest Documents