Portraits

To create a portrait online, we must first decide on the domain in which the portrait will be created and viewed. Most of the previous forms that have been proposed use the domain of home pages as a space in which people are portrayed. Since this seems like a fairly well travelled domain, as well as one that I am sure other people are going to explore, I will choose instead a different, if perhaps more tenuous domain: that of the online virtual community.

The online virtual community exists in many forms; The form I am most interested in is one based around intercommunity communication. This includes chat groups, newsgroups, person-to-person email, etc. Each of these modes of communication is part of the online community of which I speak. Indeed, the community is not restricted to any one form of communication, but rather is defined solely on the basis of the interpersonal interaction online of a group of people. Much the way that real world communities are defined by the physical interaction of people, online communities are defined by the online communicative interaction of people. A good example of this type of community is Canard. This is the domain in which I will explore the concept of portraiture.

First, we must define some of the constraints and features of such a portrait, as well as its goals, before actually designing a visual represenatation for the portrait:

• The portrait should be updated in real-time, so that viewers of the portrait can infer very current information about the subject.
• The portrait should include temporally dynamic data, such as the subject's current activity, his location, and his interactiveness, as well as the subject's temporally static data, such as age, address, and job status.
• The portrait should contain information about the interaction of the subject with other people. In this way, the portrait becomes of not just a single person, but of the surrounding group of people as well.
• The portrait should utilize all media available for its portrayal, and should represent data in its most natural format. For example, an address should appear as a location on a map, while the level of interaction with other people should be portrayed using ambient chatter sounds.
• Most importantly, the intent of the portrait should be to give the viewer a sense or feel for the subject at a given time, and should strive more for "fuzzy" portrayal of data before exact, scientific display.
• The portrait should be relatively small, so that a number of portraits can be on screen at once, and each portrait can be read (at least superficially) at a glance.

All of this information is available on the canard system, most notably the person tracking information and group interaction. What makes such information different from just online chat systems is that the information is heavily grounded in reality: the system is an extension of a real life community, and represents the interaction and presenation of people as they are in real life.

page last updated 3/27/2001