L. Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 47:107--138, 1991.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....of deliberation are in turn grounded by perceptual acts, which in negotiation is communication. 1. 3 Computational Societies Distributed Artificial Intelligence (DAI) is concerned with understanding and modeling action and knowledge in a collaborative enterprise consisting of a number of agents [20]. DAI is conceptually subdivided into two related approaches: Distributed Problem Solving (DPS) and Multi Agent Systems (MAS) 14] DPS concerns itself with the problem of how the task of solving a particular problem can be subdivided among a set of nodes that cooperate in dividing and sharing ....

....can be caused by either uncertainty in the world, different reasoning procedures or limited resources. ffl How to actually engineer and constrain practical DAI systems through the design of platforms and methodologies. As Gasser notes, the solutions to these problems are not independent [20]: different procedures for communication and interaction have implications for coordination and coherent behaviour. Different problem solving and task decompositions may yield different interactions or agent modeling requirements. Coherent, coordinated behaviour depends on how knowledge ....

L. Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 47:107--138, 1991.

....that is appropriate for their intended interactions Some of the surface issues associated with this problem are beginning to be studied in both the multiagent systems community [28] 42] and in the federated and multidatabase community. However, the deeper issues still await serious work [17]. Another long term issue is the integration of the work in the computer supported cooperative work community, the intelligent user interfaces community, and the multiagent community. If a model of computing in which networks of computational agents and 21st century, then it seems obvious that ....

# L. Gasser, "Social Conceptions of Knowledge and Action," Artificial Intelligence, vol. 47, no. 1, pp. 107--138, 1991.

....each message and action; ffl How to quickly understand the meaning of non local information. Note that the first four items suggest that learning in CDPS involves acquiring organizational design level knowledge, because they are partial solutions to the basic question which agent does what, when [9]. 4.2 Conceptual Cooperative Learning Model In this section, a model of network coordination is described in which learning is incorporated. There are three major issues involved in implementing a learning mechanism: when to invoke learning, how to coordinate learning among agents, and when and ....

L. Gasser, "Social Conceptions of Knowledge and Action," Artificial Intelligence, Vol. 47, pp. 107-138, 1991.

....options are available to them. To build such a description we must introduce the concept of agency into the model. Ours is one of the few comprehensive descriptions of computational task environments, but there are many formal and informal descriptions of the concept of agency (see [18, 19]) Rather than add our own description, we notice that these formulations define the notion of computation at one or more agents, not the environment that the agents are part of. Most formulations contain a notion of belief that can be applied to our concept of what an agent believes about its ....

Les Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 47(1):107--138, 1991.

....protocol that is appropriate for their intended interactions Some of the surface issues associated with this problem are beginning to be studied in both the multiagent systems community [28, 42] and in the federated and multi database community. However, the deeper issues still await serious work [17]. Another long term issue is the integration of the work in the computer supported cooperative work community, the intelligent user interfaces community, and the multiagent community. If a model of computing in which networks of computational agents and people seamlessly interact is to become a ....

L. Gasser, "Social Conceptions of Knowledge and Action," Artificial Intelligence, Vol. 47, No. 1, 1991, pp. 107--138.

.... to subscribe to particular agent architectures (which one would assume will be adapted to the task environment at hand) and we may ask questions about the inherent social nature of the task environment at hand (allowing that the concept of society may arise before the concept of individual agents (Gasser 1991)) Thus we might study how the concept and architecture of agents arises naturally from their task environment, rather than by starting with a pre designed agent architecture. Another example is the search for so called social laws (Shoham Tennenholtz 1992) that can be derived from a task ....

Gasser, L. (1991), `Social conceptions of knowledge and action', Artificial Intelligence 47(1), 107--138.

.... on handling the resolution of these inconsistencies as an integral part of cooperative problem solving among agents ties this work closely with recent work on negotiation in distributed problem solving [8, 9, 10] distributed consistency maintenance [11, 12, 13, 14] and open information systems [15, 16]. The remainder of the paper is broken into six sections. The next section (Section 2) first presents an example of FA C problem solving. This example provides the context for discussions on the implications of this style of agent interaction on local and cooperative problem solving. Section 3 ....

Gasser, L. "Social Conceptions of Knowledge and Action," Artificial Intelligence 47, 1991, pp. 107--138.

....(cooperative) reasons. The issue of action selection and sequencing is obviously quite complex. To further muddy the waters, interaction effects between the candidate actions preclude independent evaluation. Agents situated in large, open environments are in fact embedded in a web of influences [6] that must be considered during decision making. To illustrate, Figure 1 shows an organized network of interacting information agents in the WARREN [4] style. There are three main agent types in the network: Database Managers Agents that are experts in data maintenance and organization. These ....

Les Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 47(1):107--138, 1991.

....need to be satisfied before joint problem solving can commence and prescribes how individuals should behave once it has begun. Any theory of DAI ought to account for how aggregates of agents can achieve joint actions that are robust and continuable despite intermediate foul ups and inconsistency [12]. Responsibility offers a step towards this; providing mechanisms for controlling activity in dynamic and unpredictable environments, whilst retaining a degree of generality and predictability. Empirical evidence to substantiate this claim has been obtained [13] Compared with groups of selfish ....

L.Gasser (1991), "Social Conceptions of Knowledge and Action", Artificial Intelligence 47, 107-138

....resources 7. How to avoid or mitigate harmful overall system behavior, such as chaotic or oscillatory behavior 8. How to engineer and constrain practical MAS systems How to design technology platforms and development methodologies for MAS Solutions to these problems are of course intertwined [54]. For example, different modeling schemes for an individual agent may constrain the range of effective coordination regimes; different procedures for communication and interaction have implications for behavioral coherence; different problem and task decompositions may yield different ....

L. Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 43(1), 1991.

.... agent architecture (which one would assume will be adapted to the task environment at hand) This separation also allows a user to ask questions about the inherent social nature of the task environment at hand (allowing that the concept of society may arise before the concept of individual agents [17]) Such a conception is unique among computational approaches. ffl The representation of the task structure from three different viewpoints. The first view is a generative model of the problem solving episodes in an environment a statistical view of the task structures. The second view is an ....

Les Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 47(1):107-- 138, 1991.

....options are available to them. To build such a description we must introduce the concept of agency into the model. Ours is one of the few comprehensive descriptions of computational task environments, but there are many formal and informal descriptions of the concept of agency (see [21, 22, 27]) Rather than add our own description, we notice that these formulations define the notion of computation at one or more agents, not the environment that the agents are part of. Most formulations contain a notion of belief that can be applied to our concept of what an agent believes about its ....

Les Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 47(1):107--138, 1991.

.... to subscribe to particular agent architectures (which one would assume will be adapted to the task environment at hand) and I may ask questions about the inherent social nature of the task environment at hand (allowing that the concept of society may arise before the concept of individual agents [Gasser, 1991]) Such a conception is unique among computational approaches. I will adapt Shoham s Agent Oriented Programming terminology when I need to be more specific about an agent s internal architecture (in Chapter 5) ffl The representation of the task structure from three different viewpoints. The ....

....resources and changing agent knowledge boundaries. It is the overall collection of problem solving knowledge that is fixed not the definition of agents. Gasser and Ishida, 1991] This provides a potentially more open system and social perspective than is usual in DAI research [Hewitt, 1991, Gasser, 1991] A final problem remains with the form of the model that I can only address as a future direction : that of the mutual construction of agents and their environment. The organization of the world is not wholly dictated either by people individually or by their environments [Agre, 1991, Latour, ....

[Article contains additional citation context not shown here]

Les Gasser. Social conceptions of knowledge and action. Artificial Intelligence, 47(1):107--138, 1991.

....similar to some subset of the past problem instances. Moreover, as will be discussed below, the scope of learning encompasses activities beyond OSD. Bond and Gasser [Bond and (Eds) 1988] identify six basic problems that DAI researchers have begun to address. We will now discuss these six problems [Gasser, 1991] and indicate how learning can be of help in each of these six areas: 9 ffl How to formulate, describe, decompose, and allocate problems and synthesize results among a group of intelligent agents. Problem decomposition is the process of dividing up a problem into subproblems and allocating ....

Gasser, L. Social conceptions of knowledge and action. Artificial Intelligence, 47(1):107--138, 1991.

No context found.

Les Gasser. Social Conceptions of Knowledge and Action. Artificial Intelligence, 47, January/February, 1991.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC