Link People-to-People-to-Geographic- Place

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 
153 154 155 156 

In the above discussion we noted that proximate communities have strong

associations with physical spaces and have suggested that they would be well

supported by information systems that can tie user interactions to geographic

places, which we refer to as P3-Systems. A number of P3-Systems have been

prototyped since the early 1990s, but each has implemented only a limited set

of features. In fact, each system developed has been a narrow exploration of

a design space whose overall characteristics remain unknown. This is not

simply because of technological constraints but also because the building of

such systems has lacked a firm theoretical foundation.

A theoretical framework for understanding the utility of various designs of a

particular class of system can only come about after the class of systems under

discussion is recognized as distinct. Information systems that systematically link

people-to-people-to-geographic-place have not been considered as a related

or distinct category. This is probably on account of the dearth of location aware

systems that seamlessly link people–to-people-to-geographic-places. However

with an exponential growth in the number and types of such systems being

prototyped, it is now apparent that a new collective term is needed. Further,

while terminology such as “location based services,” “augmented reality,”

“virtual reality,” “teleportation” and “mixed reality” can be used to explain some

of the technology utilized, these terms on their own do not describe the

functional similarities of these systems. As such, we have coined the term P3-

Systems to describe various information systems that systematically link

people-to-people-to-geographic-place. P3-Systems that have been prototyped

to date have incorporated only a limited number features from the potential

design space. Recognition of P3-Systems as a distinct class of applications

allows us to distinguish between basic design features, and to provide a

theoretical/conceptual framework for future development in this area.

Traditionally, the starting point for categorizing a CMC system has been in

terms of whether it supports synchronous or asynchronous communication

(Rafaeli & Newhagen, 1996). While this categorization can be applied to many

systems, it should be noted that such categorical distinctions are not clear-cut.

Communications created using synchronous technologies can be stored and

made persistent and searchable, thus enabling asynchronous use of the medium.

Further, “synchronous communication” is not always real-time. For example,

Internet Relay Chat (IRC) requires that users hit the carriage return before the

information they have typed is shared. The result is that individuals using IRC,

and many other chat systems, can change what they were going to say before

having said it. On the other hand, asynchronous communication tools such as

email can be used for quick message exchanges that make the interactions near

synchronous. These issues highlight both the elasticity of synchronicity and the

importance of understanding the significance of making interactions persistent

(Erickson & Laff, 2001).

Synchronous and asynchronous systems can be viewed as two sides of a

continuum. Consequently, while the division into these two categories may help

us understand various design possibilities and implications, they should not be

considered absolute. Further, this categorical distinction will be extended to

include synchronous and asynchronous location awareness for systems that do

not involve traditional forms of communication. “Synchronous location awareness”

refers to the provision of current information about user location. This

location awareness need not necessarily be reciprocal, in the sense that the

system may provide a user with a buddy’s location without necessarily

providing the buddy with the user’s location. “Asynchronous location awareness”

refers to the provision of historical information about user location.

Collectively synchronous communication and synchronous location awareness

data is created with the expectation that it will be processed in near real-time,

whereas asynchronous communication and location awareness data is produced

with the expectation of an unpredictable delay between data creation

and consumption.

Beyond questions of synchronicity, existing P3-Systems primarily adopt two

basic design approaches to linking people-to-people-to-geographic-place.

First, People Centric P3-Systems are those that use user location to improve

contextualization and coordination of interactions, and to enable the identification

of previously unidentified affinities between users. People Centric P3-

Systems are people centric in the sense that the user interface provided is

focused not on a particular location but on the movements of people in physical

spaces. People’s location can be understood in both absolute and relative

terms. So a person could be located at a particular latitude, longitude and

altitude (absolute location), or the individual could be located near a friend

(relative location). The second design approach is Place Centric P3-

Systems that use virtual spaces that represent physical locations. These

systems are place centric in the sense that they use physical space/s to delineate

virtual space/s in which an associated user’s actions and interactions can be

seamlessly represented. The virtual spaces that represent physical locations

either contain online representation of people’s use of physical space or online

interactions related to physical location.

From the above discussion, we derive four basic P3-Systems design approaches:

(1) People Centric P3-Systems based on absolute user location.

(2) People Centric P3-Systems based on user co-location / proximity.

(3) Place Centric P3-Systems based on use of physical spaces by users/


(4) Place Centric P3-Systems based on interactions in virtual spaces representing

physical locations (Matching Virtual places).

Each of these four basic design approaches can be instantiated synchronously

or asynchronously; as a result there are eight distinct categories, which we

outline below. Table 1 summarizes these categories. While the design features

Table 1. P3-system features


Design Approaches

Synchronous Communication or

Synchronous Location Awareness

Asynchronous Communication or

Asynchronous Location Awareness

(1) Absolute User


Utilizes remote awareness of

current user location

Utilizes people’s location histories



(2) Colocation/


Utilizes real-time inter-user colocation

for the exchange of social


Utilizes co-location history to enable

future interactions.

(3) Use of Physical

Spaces by People

Utilizes online representation of

user’s current use of physical


Utilizes history of people’s use of a

particular space


Centric (4) Interactions

in Matching Virtual


Utilizes synchronous online

interactions spaces related to

physical location.

Utilizes asynchronous online

interactions related to physical


of most categories will be described using examples of systems identified from

the literature, a few will be described using theoretical or hypothetical systems.

People Centric P3-System

As described above, People Centric P3-Systems use community member’s

location to improve contextualization and coordination of interactions, and

enable identification of previously unidentified affinities between users. They

can be divided into two sub-categories. These are: (1) absolute user location,

i.e., based on awareness of where somebody is located; and (2) user colocation,

i.e., based on inter-user proximity. These two categories in turn will

be examined according to whether they primarily support synchronous or

asynchronous interactions/awareness.

People Centric P3-System Features Based on Absolute User Location

• Synchronous Communication or Synchronous Location Awareness

Belonging to this category is the earliest location aware P3-System,

“Active Badge,” which provides real-time information about people’s

location. Conceived, designed and prototyped between 1989 and 1992,

the Active Badge system provides a means for locating individuals within

a building by determining the location of their “Active Badge” (Want &

Hopper, 1992; Want et al., 1992). The Active Badge device worn by

personnel transmits a unique infra-red signal every 10-15 seconds that is

detected by one or more networked sensors within an equipped building.

The location of the badge (and hence its wearer) can thus be determined

on the basis of information provided by these sensors. System users use

the command — FIND (name) — that provides the current location of the

named badge, and a list of all the locations it has been sighted at in the last

five minutes. The system is designed to coordinate communication between

individuals. For example, internal and external phone calls could be

routed to the phone nearest to the location of an individual based on the

location of the Active Badge bearer. The Active Badge system has an

online community space that could be used to locate employees without

a public-address system or without telephoning all the possible locations

at which they might be found. Although the system was not designed with

the aim of linking people-to-people-to-geographic-place, by providing

awareness of people’s movements in remote locations, this system

enables such links.

It is not hard to imagine various applications that can be supported by the

remote locatability feature of the Active Badge system. For example, a

location-aware descriptor could be provided next to an instant messaging

buddy. In fact, commercial services that allow for the remote location of

mobile phone users are already in use. For example Ulocate (http:// allows users to see the location of all family members

using the system displayed on a map on a 24/7 basis.

An alternative to the above approach is to reverse the process, so that

instead of a person seeking out the location of somebody they wish to

communicate with, the system provides details of the location of the

originator of an incoming communication. For example, the caller-ID of

incoming phone calls could also contain a location descriptor, or the

message lines in a private online chat could contain details of the location

of the sender. This approach is used by the ActiveCampus Explorer,

where users can have the system automatically index their instant messaging

messages with a descriptor of their current location (Griswold et al.,


• Asynchronous Communication or Asynchronous Location Awareness

Systems under this category provide asynchronous location awareness.

Group calendars that systematically describe the location of individuals

over time fall under this category. Systems that provide details of the

location of the originator of an incoming asynchronous communication

also fit here. For example, an email could be indexed by the location of the

sender. In fact, some moblogs (mobile phone web logs) contain pictures

that are time and location stamped. In other words, by linking location

awareness to asynchronous communication these systems render the

location awareness asynchronous.

People Centric P3-System Features Based on Co-Location/Proximity

• Synchronous Communication or Synchronous Location Awareness

This category of people centric P3-Systems uses inter-user proximity or

co-location to electronically collect or manage the exchange of social

information. A simple application of this approach is chat between co228

located individuals, such as that provided by Cybiko (

a child’s toy that uses Radio Frequency communication. Cybiko allows

ad hoc networks to be formed between individuals that are within a short

distance so that they can interact electronically. Similarly, the Hummingbird,

a small portable device, supports social awareness between people

who are co-located (Weilenmann & Holmquist, 2001). The Hummingbird

uses wireless communication to give members of a group continuous

auditory and visual indications of other group members in the vicinity. The

designers of Hummingbird hoped to support face-to-face interactions by

visualizing group member proximity.

Another system that provides support for face-to-face interactions using

co-location is Borovoy et al.’s (1998) “Meme Tags and Community

Mirrors” System. Meme tags are a class of groupware tags designed to

build, in the authors’ own words, “community” (p.159). The system has

both a personalized online space (individual Meme-Tags) and a community

space (Community Mirrors). In this system a meme is an idea or

opinion, expressed as a short piece of text. The Meme Tag contains

community relevant memes that a participant has chosen. Inter-user

proximity enables the spread of memes from person to person synchronously.

This is meant to encourage people-to-people interactions. The

purpose of the Community Mirrors is to convey a variety of information

about meme exchanges between users in near real-time to other users.

Included within these displays are the actual texts of the memes, popular

ideas, dying ideas, as well as information about group dynamics, such as

the “cliquishness” of the gathering. These Community Mirrors also give

users a sense of what other participants know. The system design aims to

facilitate the formative stages of interaction by providing people with

additional common reference points for conversation.

Co-location can also be used specifically for social matching. To date, the

social matching has typically been in terms of supporting dating rather than

proximate community, but the design approach is of potential value in both

situations. A system that illustrates this basic design idea is the Japanese

dating toy LoveGety (Reuters/Wired News, 1998). When a blue (male)

LoveGety and a pink (female) LoveGety are within 15 feet of one

another, they beep and flash, telling the user that another LoveGety owner

is close by. Codes such as “talk,” “karaoke,” and “get2,” with a variety

of meanings are used to communicate what the user is interested in. This

system design encourages real-time face-to-face interactions around the

idea of “dating,” however the proximity approach can logically be used to

encourage interactions based on many other types of social matches.

Another, example of a social matching system is Proxy Lady, a mobile

system for informal, opportune face-to-face communication, running on a

PDA equipped with a radio transceiver (Dahlberg et al., 2000). Proxy

Lady lets the user associate information items (e.g., emails) with other

people, called “candidates for interaction.” When a “candidate for

interaction” is in the proximity, Proxy Lady notifies and provides the user

with the associated information item (e.g., the email message), and if

suitable, is followed by an informal face-to-face interaction.

“Social Net” infers interest matching from patterns of co-location over

time to recognize social relationships and infer affinities between users

(Terry et al., 2002). The Social Net handheld client records the time and

duration of physical co-location synchronously, and searches for recurring

patterns of co-location to asynchronously infer shared interests

between users. The system contains two lists, a “friends list” that contains

users who are friends, and an “unknown list” that contains users who are

not friends that one comes in proximity or contact with in some sort of

consistent pattern. This is achieved by recording of encounters between

co-located users, as well as their time and duration, in an encounter

record. Periodically, the encounter record is examined and if a suitable

pattern of co-location is observed between individuals, the name is

included in the unknown list. While the friends list is manually created by

the user, the unknown list is automatically and synchronously generated by

the system and not visible to the user due to privacy concerns. When two

“friends” come in contact with each other, their systems communicate by

comparing their “unknown lists.” On detection of a friend by one system

in the other system’s unknown list, the system informs the user of a

potential new friend recommendation that can be made.

In addition to social matching, proximity can also be used to enable or

support synchronous information exchanges, which in turn is often supported

by various asynchronous components. For example, Hocman, a

mobile peer-to-peer application, supports social interaction between

motorcyclists (Esbjörnsson et al., 2003). Hocman users provide personal

information of themselves and their bikes in HTML pages, which is

exchanged with other bikers equipped with a Hocman, typically at traffic

light stops. This synchronous interaction is accompanied by audio notifications.

There is also a major asynchronous component to Hocman,

which will be discussed in the section below. Similarly, RoamWare

(Wiberg, 2001) uses proximity to semi-automatically identify when individuals

get together for ad-hoc meetings, and then supports synchronous

ad hoc mobile meeting note taking. Mobile meeting notes can then be

shared asynchronously using a suite of CMC tools once users return to

their desktop computers. Finally, FolkMusic (Wiberg, 2004) uses

proximity to trigger services for music sharing between co-located

individuals. The fully instantiated system will also use GPS receivers to

map audio traces left by individuals to geographic locations, resulting in

music files being associated with specific locations.

• Asynchronous Communication or Asynchronous Location Awareness

To date, no P3-systems based on user proximity that we know of provide

an interaction/communication framework that is primarily asynchronous.

RoamWare with its focus on seamlessly connecting planned and mobile

ad hoc meeting through the distribution of mobile meeting notes using

synchronous chat and asynchronous email through the RoamWare Desktop

makes aspects of the ad hoc meetings asynchronous. As mentioned

above, Hocman, the application that supports social interaction between

motorcyclists has a significant asynchronous component. That is, the

system is designed so that when the biker ends her/his journey, s/he

browses the pages received from other motorcyclists. Social Net also

allows for asynchronous exchanges between users, by waiting for friends

to interact before exchanging information about unknown but potential


It is possible to imagine numerous designs of hypothetical systems that

would fit into this category by more extensively utilizing asynchronicity.

Some possible systems are those that stretch our understanding of

proximity to include asynchronous use of shared physical space rather

than synchronous use of physical space. This would be similar to Social

Net but would require a history of a user’s use of various geographical

locations. This would allow the identification of affinities such as similar

routes of travel to work, use of gyms, etc., even if they occurred at

different times of the day. This in turn could be used to encourage faceto-

face interactions when individuals are co-located synchronously. This

design would still be people centric because it would be based on

individual location history data, however, it should be noted that this is

perhaps most easily achieved by taking an absolute rather than relative

proximity approach to user location. An alternative approach is to

asynchronously use, at a later, more convenient time, the data recorded

synchronously by the system when individuals were co-located in real

time. While this is done by Hocman, the early prototypes described in the

literature did not strongly support interaction post data exchange. A

stronger example might be a system that automatically exchanges

individual’s business cards electronically when co-located, and then

encourages asynchronous interactions through the provision of personal

card-exchange histories tied to a social network visualization and an

asynchronous communication system such as email.