Since the agent paradigm emerged, agent researchers have faced the challenge of build-ing open societies in which heterogeneous and independently designed entities can work towards similar or different ends. Open societies involve agents that do not necessarily share the same interests, that do not know and might not trust each other, but that can work together and help each other. One of the key omissions in the computational rep-resentation of open societies relates to the need for norms in multi-agent systems, that help to cope with the heterogeneity, the autonomy and the diversity of interests among their members. This also requires agents that can reason about norms because their par-ticipation in a society, rather than predefined, must be voluntary. So, these agents must understand why norms should be adopted and complied with, and why the authority and the power of agents in a society must be respected. This thesis addresses both the in-troduction of norms in systems of autonomous agents, and the modelling of agents that can reason about norms. The thesis makes three main contributions. First, it develops a framework of norma-tive concepts that enables agents to reason about norms and the society in which they participate. Second, it provides the means for agents to identify situations of power, and to use these powers both for the satisfaction of their goals and to understand why the goals of other agents must be satisfied. This is required since agents in an open soci-ety must interact with other agents which are also autonomous, and power represents a means to influence them. Third, this thesis provides models for agents that adopt and comply with norms not as an end, but as the result of a deliberation process in which their goals and motivations are taken into account. This enables agents to voluntarily decide whether participating in a society is important for the achievement of their goals.

o relevant agent literature, without claiming these to be either exhaustive or the most authoritative references. Scale: It is almost a platitude to mention the size of the Internet in general, and of the Web in particular. Google now indexes well over a billion pages, and the number of connected hosts runs into the millions. These numbers are orders of magnitude larger than any traditional single knowledge-base for which much of current KR technology has been designed [Turner and Jennings, 2000]. Change rate: Many portions of the Internet display a very high change rate, with information changing on the timescale of days (e.g. news sites), hours (e.g. auction sites), or even minutes (e.g. stockmarkets). On the other hand, KR techniques, such as those from knowledge engineering, typically have been designed for update rates in the order of months, or even slower. (e.g. [Schut and Wooldridge, 2000]) Lack of referential integrity: One of the major departures that the Web took from trad

. A model of BDI--agent in game--theoretic framework is presented. The desire is represented as agent's goal to achieve a maximum level of utility. A reasoning process based on agent's rational behavior is proposed. This process determines agent's intention. It is also shown how to use the backward induction consistently with the assumption of the common knowledge of rationality. 1 Introduction We are going to discuss the following problem: How does a rational agent use its knowledge in decision making ? Since the problem is general, we put it in a game--theoretic framework. In the theory of games, agent's rationality is understood as a way of maximizing the utility of the agent relatively to its knowledge. The knowledge may concern the game that is to be played as well as the agents participating in a play. The main task of the paper is to model BDI-agent that is supposed to live in the world of dynamic games. Agent's belief is identified with the knowledge about the game and abo...

Coordination is a crucial problem in CE systems and it is neither easy to obtain nor to maintain. Our work is an attempt to develop a general model for coordination which can be adapted for some situations in the context of CE. For this purpose, the coordination definition developed by Malone [25] has been adopted. Coordination is then de#ned as the process of managing dependencies between activities. In this context, a theoretical model is presented that allows one to determine how to model an agent's activities and how to detect dependencies between those activities. In our model, major concepts are developed in terms of components of coordination, situations of coordination, coordination mechanisms and the coordination process. In this paper, we detail this model and then, we present an illustrative example and finally, we identify the current status and the future evolution of our approach.

Abstract. The aim of this chapter is to discuss the applicability of recently proposed knowledge modelling tools to the development of agent-based systems. The discussion is derived from the real world experience of a particular software tool

A belief revision procedure is composed of four phases: detection of an inconsistency, identification of the culprit(s), decision of a context to be maintained and propagation of the chosen context in the belief base. In this paper, we present a model and an implementation for the decision phase of such a procedure, to be used in an open multi-agent context. By open MAS, we denote a system where agents may dynamically enter and leave the agency whenever they want, and where they cooperate with each other by proposing and accepting to form coalitions in order to achieve one of their goals. Agents must have therefore a social reasoning mechanism, which enables them to reason about the others. We consider that the procedure of coalition formation --- if not well succeeded --- may lead an agent to revise his beliefs about the others. This process is done autonomously, i.e., without any pre-established centralized control. Our model is based on the notions of information source and information topic. We have implemented this model within the agents of the DEPINT system. We present the simulation results of a running session of the system, The work described in this paper was partially developed during the author's PhD program at LIFIA laboratory, Grenoble, France, when he was supported by FAPESP, grant number 91/1943-5. The author would like also to thank Rosaria Conte and Cristiano Castelfranchi, IP/CNR, Rome, Italy, for the useful discussions carried on during this period.

We define the notion of social conventions in a standard gametheoretic framework, and identify various criteria of consistency of such conventions with the principle of individual rationality. We then investigate the emergence of such conventions in a stochastic setting; we do so within a stylized framework currently popular in economic circles, namely that of stochastic games. This framework comes in several forms; in our setting agents interact with each other through a random process, and accumulate information about the system. As they do so, they continually reevaluate their current choice of strategy in light of the accumulated information. We introduce a simple and natural strategy-selection rule, called highest cumulative reward (HCR). We show a class of games in which HCR guarantees eventual convergence to a rationally acceptable social convention. Most importantly, we investigate the efficiency with which such social conventions are achieved. We give an analytic lower bound ...

This paper investigates the use of multi-agent systems (MAS) for natural language

At the heart of cooperation between agents we can find the notion of agent modeling. The representation one agent maintains of other agents is indeed essential for social reasoning. In order to cooperate, an agent must know how it may cooperate with each of its acquaintances and such knowledge is tightly connected to the known characteristics of another agent. Agent modeling is even more significant if we consider open systems. In such systems, agents have to update their representation of others in order to learn a description of any incoming agent. In this paper, we propose a model of representation of others based on the external description defined by Sichman. We make use of our model to allow incoming agents to introduce themselves to others and to classify their acquaintances into stereotypes. These two abilities permit us to conceive an open multi-agent system (MAS) keeping the agent integration task distributed among the agents. An application of such a MAS to the building of open distributed information systems is presented.