Takeshi Ohguro, Sen Yoshida, and Kazuhiro Kuwabara
NTT Communication Science Laboratories,
2-4 Hikaridai, Seika-cho, Soraku-gun, Kyoto 619-0237 JAPAN
{ohguro,yoshida,kuwabara}@cslab.kecl.ntt.co.jp
Recently, implicit, unspoken knowledge, experience and memory - sometimes captured as ``tacit knowledge'' - is considered important and getting growing attention. Unfortunately, current network communication tools are mostly limited to treat explicit form of knowledge only, thus supports for conveying and sharing tacit knowledge are still weak. In this paper we propose a system for convey (and share) the presence and status information of people. Such information is considered as the basis for tacit knowledge which needs to be shared in the community.
As the information technology such as the Internet grows, network communications become important factors in our daily lives. Needless to say, use of e-mail and Web access become already popular. Even for mobile telecommunication devices (such as a cellular phone service and a personal handyphone system: PHS), there are additional communication services which are linked with the Internet: For example, ``short mail service'' (between PHSs) can be gatewayed to the Internet mail.
While these tools are very convenient, they are not completely satisfactory: For example there are problems such as information overload, which beat us with stress sometimes. Another source of problem is that these tools are mainly oriented to written or spoken languages: That is, these tools are mostly limited to treat explicit form of knowledge only. The problem is well illustrated by the ``flame wars,'' which is very common in Usenet news, mailing-lists or chat boards [1]. The main cause of ``flame wars'' seems to be the lack of contexts (the depth of knowledge of individuals, or their attitudes towards the subjects), ethics (mutual knowledge and manners for network communication), mood behind the statements, and so on. Such information as contexts and ethics is usually not expressed as explicit forms, instead they are shared as implicit knowledge.
Such an implicit, unspoken knowledge, experience and memory is considered important and getting growing attention recently, typically in areas of knowledge management and organizational learning. The concept is captured as the ``tacit knowledge,'' which is contrasted to the ``explicit knowledge'' [2]. The importance of tacit knowledge and its interaction with explicit knowledge is argued in [3]. Furthermore it is pointed out that there are four processes in knowledge conversion, according to the types of source knowledge and converted knowledge. Among them, we focus on the ``socialization'' process, which is a process that convert and convey (one's) tacit knowledge to tacit knowledge (of others).
There are several researches that can be considered as supporting the ``socialization'' process over the network. One direction is that researches and developments for ``multimedia'' applications. Utilizing multimodal interfaces, nonverbal information which may contain tacit knowledge can be conveyed. Socia [4] and FreeWalk [5] are the good examples along with the line. They not only provide multimodal information channels but also try to support the awareness in communications. While these applications are powerful in multimodal communication, these are not ultimate solutions: For example, many users of TV conference system complain about the lack of liveness and feel they are ``separated into both sides.'' Furthermore, these system require high bandwidth.
Another direction is to externalize one's memory and share these informations. In the Knowledgeable media project [6], the concept of ``Knowledge as Media,'' which seems to be closely related to the notion of tacit knowledge, is proposed. The concept says that knowledge lies in interactions between people, not in individual people in closed form. Along with the concept, CoMeMo-Community system is being developed [7]. In the system, personal associative memory is externalized, shared and being interacted each other.
There is still another direction that extract and make use of physical data of the users. An MIT group starts studying Affective computing (1) which tries to extract emotions and affect signals from biological and physiological data of the users, and to make use of these emotions and affect signals in supporting communications and several other computing. Ishii's Tangible bits research project (2) also makes use of physical (but not physiological) data as core information to be conveyed [8]. Especially inTouch [9] explores new forms of interpersonal communication through touch, via the movement of ``shared'' object which can be touched and moved by the users who are geographically distributed.
Along with these lines, we are considering another communication tool that convey (and share) tacit knowledge over the network. As the information which should be conveyed, we choose the presence and status information of people. That is, information such as ``I'm here'' and ``Are you there?'' or, ``Who's caring me'' and ``I care you.'' Especially in a networked environment, such information as presence and status tends to remain unknown, because:
Motivated by the reasons above, we propose a system called Gleams of people, which provides the presence and status information of people. The system is designed to be an multi-agent system, which provides very simple, intuitional messaging (signaling), with easy and handy interface.
The function of Gleams of people is similar to ``Who's online,'' which is often provided by online community services and applications such as ICQ (3). In such ``Who's online'' services, current status of users are provided, such as online/offline, busy/left the seat/idle, and accept/reject. These information are typically used to decide whether to start communication or not. That is, users are expected to start communication by other methods when using the service. On the other hand, we do not assume that communication will be started after using our presence monitor. Thus, our system is conceptually very different to these services: To provide presence and status information itself, as a very simple, intuitional message, is the aim of our system.
There exist some services on the net that provide personal information such as whois (4) and Bigfoot (5). Several portal sites also provide ``person finder.'' However these services are rather static (and often out-dated) and status information for each person is missing in general.
Intuitively Gleams of people functions like a /sbin/ping (6) among people (more precisely, among personal agents). Unlike the ordinary ping, ``ping'' message in Gleams of people carries two additional informations. One is a ``color'', which represents current ``mood'' of the sender. The other is a ``level'', which indicates whether the ping is done by the system automatically or by the user's intention.
Though the presence and status information is intended to be conveyed and
shared,
the motivation for sending and receiving such information should be
very personal one for individual people.
Hence the system does not fit in conventional server-client architecture,
we need to adopt multi-agent architecture for
The system consists of personal agents for individual users and a (cluster of) repeater agent(s). Each personal agent maintains sets of friends of the user, mediates the transmission of presence and status information in place of the user, and displays these status for the user. A repeater agent can be considered as a shared ``buffered repeater.'' It is used by personal agents for store and forward information when the destination agent cannot be found or is offline. The repeater agent is needed since information while the user is offline is still valuable as a presence and status informations of others, if they are not too out-dated. Moreover, the fact that transmission is not always immediate gives users some relief that communication in this system is not so disturbing one.
From user's point of view, the system works as follows:
Fig. 1. planned screen image of the system
Figure 1 is the planned screen image of the system. Light circles represent the members of current friend set. Each circle gleams when a ping from the friend is received, with the color of the ping (or when a response to the ping by the user is returned). Furthermore it gleams in a certain interval, which is computed from past record (frequency, direction, and time, etc.) of pings. A circle gleams stronger for level two ping, weaker for level one ping. While the difference between ping level is that whether it is system initiated or user initiated, users might find another meaning for ping levels. That is, users can naturally interpret the ping as ``I'm here'' and ``Are you there?'' for level one ping, ``He/She is caring me'' and ``I care you'' for level two ping.
In this paper we proposed the concept and the design of the system for presence and status monitoring of people, Gleams of people. The presence and status information of people is considered as the basis for tacit knowledge, which needs to be conveyed and shared in the community. Systems of such a direction will play important roles in future network communications.
This is a first step for us that tries to convey tacit knowledge over the network. From the point of view of community knowledge, the support via proposed system is still limited to a personal level support, rather than a community level support. Future directions for the community level support are:
Further issues to be studied include: