Skip to main content
Fachgebiete
Automobil + Motoren Bauwesen + Immobilien Business IT + Informatik Elektrotechnik + Elektronik Energie + Nachhaltigkeit Finance + Banking Management + Führung Marketing + Vertrieb Maschinenbau + Werkstoffe Versicherung + Risiko
DE
EN
Themenseiten
Organisationspsychologie Projektmanagement Marketing Smart Manufacturing
Bücher
Zeitschriften
Weitere Formate
Podcasts Webinare Technik Webinare Wirtschaft Kongresse Veranstaltungskalender Awards
Jetzt Einzelzugang starten
Zugang für Unternehmen
Referenzkunden
MTZ-Fachkongress

Antriebe und Energiesysteme von morgen


Austausch von Automobil- und Energiewirtschaft

Am 14./15. Mai geht es in Chemnitz um die Mobilitätswende durch Elektro- und Wasserstofffahrzeuge.
Erweiterte Suche
Anmelden
nach oben

2000 | Buch

Reasoning about Visibility, Perception, and Knowledge Kapitel lesen Erstes Kapitel lesen

Intelligent Agents VI. Agent Theories, Architectures, and Languages

6th International Workshop, ATAL’99, Orlando, Florida, USA, July 15-17, 1999. Proceedings

herausgegeben von: Nicholas R. Jennings, Yves Lespérance

Verlag: Springer Berlin Heidelberg

Buchreihe : Lecture Notes in Computer Science

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten
insite
SUCHEN

Über dieses Buch

Intelligent agents are one of the most important developments in computer science in the 1990s. Agents are of interest in many important application areas, ranging from human-computer interaction to industrial process control. The ATAL workshop series aims to bring together researchers interested in the core aspects of agent technology. Speci?cally, ATAL addresses issues such as th- ries of agency, software architectures for intelligent agents, methodologies and programming languages for realizing agents, and software tools for developing and evaluating agent systems. One of the strengths of the ATAL workshop series is its emphasis on the synergies between theories, infrastructures, architectures, methodologies, formal methods, and languages. This year’s workshop continued the ATAL trend of attracting a large n- ber of high-quality submissions. In more detail, 75 papers were submitted to the ATAL-99 workshop, from 19 countries. After stringent reviewing, 22 papers wereacceptedforpresentationattheworkshop.Aftertheworkshop,thesepapers were revised on the basis of comments received both from the original reviewers and from discussions at the workshop itself. This volume contains these revised papers.

Inhaltsverzeichnis

Frontmatter

Section I: Agent Theories

Reasoning about Visibility, Perception, and Knowledge
Abstract
Although many formalisms have been proposed for reasoning about intelligent agents, few of these have been semantically grounded in a concrete computational model. This paper presents \(\cal VSK\) logic, a formalism for reasoning about multi-agent systems, in which the semantics are grounded in an general, finite state machine-like model of agency. \(\cal VSK\) logic allows us to represent: what is objectively true of the environment; what is visible, or knowable about the environment; what the agent perceives of the environment; and finally, what the agent actually knows about the environment. \(\mathcal{VSK}\) logic is an extension of modal epistemic logic. The possible relationships between what is true, visible, perceived, and known are discussed and characterised in terms of the architectural properties of agents that they represent. Some conclusions and issues are then discussed.
Michael Wooldridge, Alessio Lomuscio
A Spectrum of Modes of Knowledge Sharing between Agents
Abstract
The logic S5n is widely used as the logic of knowledge for ideal agents in a multi-agent system. Some extensions of S5n have been proposed for expressing knowledge sharing between the agents, but no systematic exploration of the possibilities has taken place. In this paper we present a spectrum of degrees of knowledge sharing by examining and classifying axioms expressing the sharing. We present completeness results and a diagram showing the relations between some of the principal extensions of S52 and discuss their usefulness. The paper considers the case of a group of two agents of knowledge.
Alessio Lomuscio, Mark Ryan
Observability-based Nested Belief Computation for Multiagent Systems and Its Formalization
Abstract
Some agent architectures employ mental states such as belief, desire, goal, and intention. We also know that one often has a belief about someone else’s belief (nested belief), and one’s action is decided based on the nested belief. However, to the best of our knowledge, there is no concrete agent architecture that employs nested beliefs for decision. The reason is simple: we do not have a good model of nested belief change. Hence, interesting technological questions are whether such a model can be devised or not, how it can be implemented, and how it can be used. In a previous paper, we proposed an algorithm for nested beliefs based on observability and logically characterized its output. Here, we propose another algorithm with improved expressiveness and efficiency.
Hideki Isozaki, Hirofumi Katsuno
On the Correctness of PRS Agent Programs
Abstract
Although software agents are becoming more widely used, methodology for constructing agent programs is poorly understood. In this paper, we take a step towards specifying and proving correctness for a class of agent programs based on the PRS architecture, Georgeff and Lansky [9], one of the most widely used in industrial settings. We view PRS as a simplified operating system capable of concurrently running a series of plans, each of which at any time is in a state of partial execution. The PRS system is construed as using a simplified interrupt mechanism to enable it, using information about goal priorities, to “recover” from various contingencies so that blocked plans can be resumed and eventually successfully completed. We develop a simple methodology for PRS program construction, then present a formalism combining dynamic logic and context-based reasoning that can be used to reason about such PRS plans.
Wayne Wobcke
Incorporating Uncertainty in Agent Commitments
Abstract
Commitments play a central role in multi-agent coordination. However, they are inherently uncertain and it is important to take these uncertainties into account during planning and scheduling. This paper addresses the problem of handling the uncertainty in commitments. We propose a new model of commitment that incorporates the uncertainty, the use of contingency analysis to reduce the uncertainty, and a negotiation framework for handling commitments with uncertainty.
Ping Xuan, Victor R. Lesser

Section II: Agent and System Architectures

Rational Cognition in OSCAR
Abstract
Stuart Russell [14] describes rational agents as “those that do the right thing”. The problem of designing a rational agent then becomes the problem of figuring out what the right thing is. There are two approaches to the latter problem, depending upon the kind of agent we want to build. On the one hand, anthropomorphic agents are those that can help human beings rather directly in their intellectual endeavors. These endeavors consist of decision making and data processing. An agent that can help humans in these enterprises must make decisions and draw conclusions that are rational by human standards of rationality. Anthropomorphic agents can be contrasted with goal-oriented agents — those that can carry out certain narrowly-defined tasks in the world. Here the objective is to get the job done, and it makes little difference how the agent achieves its design goal.
John L. Pollock
Agents for Information Broadcasting
Abstract
In this paper we consider an environment which consists of one broadcasting entity (producer) which broadcasts information to a large number of personal computer users, who can down-load information to their PC disks (consumers). We concentrate on the most critical phase of the broadcasting system operation, which is the characterization of the users’ needs in order to maximize the efficiency of the broadcast information. Since the broadcasting system can not consider each user in isolation, it has to consider certain communities of users. We have proposed using a hierarchic distributed model of software agents to facilitate receiving feedback from the users by the broadcasting system. These agents cluster the system’s users into communities with similar interest domains. Subsequently, these agents calculate a representative profile for each community. Finally, the broadcasting agent builds an appropriate broadcasting program for each community. We developed a simulation of the broadcasting environment in order to evaluate and analyze the performance of our proposed model and techniques. The simulation results support our hypothesis that our techniques provide broadcasting programs, which are of great interest to the users.
Esther David, Sarit Kraus
On the Evaluation of Agent Architectures
Abstract
By now, intelligent agents have been on the research agenda of the computer science community for roughly one decade. Still, control architectures for autonomous intelligent systems have been an important research issue for an even much longer time, going as far back as James Watt’s steam engine control mechanism based on mechanical feedback. More recent work includes the development of mathematical models for control in the field of cybernetics (most notably, Wiener). Also, arguably one of the biggest contributions of more than forty years of research in Artificial Intelligence were methods and architectures aiming to describe, control, and adopt intelligent autonomous systems.
Henry Hexmoor, Marcus Huber, Jörg P. Müller, John Pollock, Donald Steiner
Toward a Methodology for AI Architecture Evaluation: Comparing Soar and CLIPS
Abstract
We propose a methodology that can be used to compare and evaluate Artificial Intelligence architectures and is motivated by fundamental properties required by general intelligent systems. We examine an initial application of this method used to compare Soar and CLIPS in two simple domains. Results gathered from our tests indicate both qualitative and quantitative differences in these architectures and are used to explore how aspects of the architectures may affect the agent design process and the performance of agents implemented within each architecture.
Scott A. Wallace, John E. Laird
Reactive-System Approaches to Agent Architectures
Abstract
We present a reactive-system view to describe agents for real-world applications operating in complex, dynamic, and nondeterministic environments. We first identify agent tasks and environments to highlight the desired features in agent architectures for the tasks and environments. We then compare various architectures according to the identified features.
Jaeho Lee
A Planning Component for RETSINA Agents
Abstract
In the RETSINA multi-agent system, each agent is provided with an internal planning component—HITaP. Each agent, using its internal planner, formulates detailed plans and executes them to achieve local and global goals. Knowledge of the domain is distributed among the agents, therefore each agent has only partial knowledge of the state of the world. Furthermore, the domain changes dynamically, therefore the knowledge available might become obsolete.
To deal with these issues, each agent’s planner allows it to interleave planning and execution of information gathering actions, to overcome its partial knowledge of the domain and acquire information needed to complete and execute its plans. Information necessary for an agent’s local plan can be acquired through cooperation by the local planner firing queries to other agents and monitoring for their results. In addition, the local planner deals with the dynamism of the domain by monitoring it to detect changes that can affect plan construction and execution. Teams of agents, each of which incorporates a local RETSINA planner have been implemented. These agents cooperate to solve problems in different domains that range from portfolio management to command and control decision support systems.
Massimo Paolucci, Onn Shehory, Katia Sycara, Dirk Kalp, Anandeep Pannu
A Scalable Agent Location Mechanism
Abstract
Large scale open multi-agent systems where agents need services of other agents but may not know their contact information require agent location mechanisms. Solutions to this problem are usually based on middle-ware such as matchmakers, brokers, yellow-pages agents and other middle agents. The disadvantage of these is that they impose infrastructure, protocol and communication overheads, and they do not easily scale up. We suggest a new approach to agent location, which does not require middle agents and protocols for using them. Our approach is simple and scales up with no infrastructure or protocol overheads, thus may be very useful for large scale MAS. In this paper, we analytically study the properties of our approach and discuss its advantages.
Onn Shehory

Section III: Agent Languages

Reactivity in a Logic-Based Robot Programming Framework
Abstract
A robot must often react to events in its environment and exceptional conditions by suspending or abandoning its current plan and selecting a new plan that is an appropriate response to the event. This paper describes how high-level controllers for robots that are reactive in this sense can conveniently be implemented in ConGolog, a new logic-based agent/robot programming language. Reactivity is achieved by exploiting ConGolog’s prioritized concurrent processes and interrupts facilities. The language also provides nondeterministic constructs that support a form of planning. Program execution relies on a declarative domain theory to model the state of the robot and its environment. The approach is illustrated with a mail delivery application.
Yves Lespérance, Kenneth Tam, Michael Jenkin
Extending ConGolog to Allow Partial Ordering
Abstract
In this paper we extend the high level execution language ConGolog (developed at the University of Toronto) by adding to it a new construct which we call the htn-construct. The new construct improves ConGolog by allowing easy specification of non-determinism when a partial ordering between a set of actions needs to be maintained. Furthermore, it allows temporal constraints to be specified easily. We present an implementation of the htn-construct in PROLOG which can be directly added to PROLOG implementations of ConGolog interpreters.
Chitta Baral, Tran Cao Son
Operational Semantics of Multi-Agent Organizations
Abstract
This paper introduces a formal description of the operational semantics of multiagent organizations expressed in the Aalaadin generic model. This formalization is based on the π-calculus and the Chemical Abstract Machine (Cham).
By mapping an agent to a set of π-calculus processes and groups to Cham solutions, we show that it is possible to associate a precise semantics for the definition and dynamics of agents, groups and roles, independently of any implementation.
Our show that formalization verifies the properties of Aalaadin: agents act in several groups simultaneously, communications are described through abstract roles interaction, and organization management is performed by agents.
Jacques Ferber, Olivier Gutknecht
Open Multi-Agent Systems: Agent Communication and Integration
Abstract
In this paper, we study the open-ended nature of multi-agent systems, which refers to the property of allowing the dynamic integration of new agents into an existing system. In particular, the focus of this study is on the issues of agent communication and integration. We define an abstract programming language for open multi-agent systems that is based on concepts and mechanisms as introduced and studied in concurrency theory. Moreover, an important ingredient is the generalisation of the traditional concept of value-passing to a communication mechanism that allows for the exchange of information. Additionally, an operational model for the language is given in terms of a transition system, which allows the formal derivation of computations.
Rogier M. van Eijk, Frank S. de Boer, Wiebe van der Hoek, John-Jules C. Meyer
Toward Team-Oriented Programming
Abstract
The promise of agent-based systems is leading towards the development of autonomous, heterogeneous agents, designed by a variety of research/industrial groups and distributed over a variety of platforms and environments. Teamwork among these heterogeneous agents is critical in realizing the full potential of these systems and scaling up to the demands of large-scale applications. Unfortunately, development of robust, flexible agent teams is currently extremely difficult. This paper focuses on significantly accelerating the process of building such teams using a simplified, abstract framework called team-oriented programming (TOP). In TOP, a programmer specifies an agent organization hierarchy and the team tasks for the organization to perform, abstracting away from the innumerable coordination plans potentially necessary to ensure robust and flexible team operation. Our TEAMCORE system supports TOP through a distributed, domain-independent layer that integrates core teamwork coordination and communication capabilities. We have recently used TOP to integrate a diverse team of heterogeneous distributed agents in performing a complex task. We outline the current state of our TOP implementation and the outstanding issues in developing such a framework.
David V. Pynadath, Milind Tambe, Nicolas Chauvat, Lawrence Cavedon

Section IV: Agent-Oriented Software Engineering

Agent-Oriented Software Engineering
Abstract
The ATAL workshops focus on the links between the theory and practice of intelligent agents. One aspect of this, which is steadily growing in importance, is the idea of agent technology as a software engineering paradigm. Previous ATAL workshops have had special tracks on programming languages for agent-oriented development, and methodologies for agent system development. ATAL-99 aims to build on this experience by focussing on the wider issues of agents as a software engineering paradigm.
Stefan Bussmann, Paolo Ciancarini, Keith Decker, Michael Huhns, Michael Wooldridge
Multiagent System Engineering: The Coordination Viewpoint
Abstract
The paper focuses on the design of multiagent systems and argues that traditional approaches fall short when dealing with complex multiagent systems. On this basis, this paper shows how an approach based on coordination models can help in the design of multiagent systems. A simple example in the area of conference management is assumed as a case study to clarify the concepts expressed.
Paolo Ciancarini, Andrea Omicini, Franco Zambonelli
Using Multi-context Systems to Engineer Executable Agents
Abstract
In the area of agent-based computing there are many proposals for specific system architectures, and a number of proposals for general approaches to building agents. As yet, however, there are comparatively few attempts to relate these together, and even fewer attempts to provide methodologies which relate designs to architectures and then to executable agents. This paper provides a first attempt to address this shortcoming; we propose a general method of defining architectures for logic-based agents which can be directly executed. Our approach is based upon the use of multi-context systems and we illustrate its use through the specification of a simple agent.
Jordi Sabater, Carles Sierra, Simon Parsons, Nicholas R. Jennings
Structuring BDI Agents in Functional Clusters
Abstract
The development of complex agents requires adequate conceptual and software tools that allow modular development and software reuse. We present a concept, called capability , which represents a cluster of components of a BDI agent. Capabilities encapsulate beliefs, events and plans while, at the same time, allowing global meta-level reasoning. Capabilities enable software reuse, and are well suited as building blocks for the development of multi-agent systems. We present an implementation of capabilities within the commercial Java-based multi-agent framework JACK Intelligent AgentsTM.
Paolo Busetta, Nicholas Howden, Ralph Rönnquist, Andrew Hodgson
Towards a Distributed, Environment-Centered Agent Framework
Abstract
This paper will discuss the internal architecture for an agent framework called DECAF (Distributed Environment Centered Agent Framework). DECAF is a software toolkit for the rapid design, development, and execution of “intelligent” agents to achieve solutions in complex software systems. From a research community perspective, DECAF provides a modular platform for evaluating and disseminating results in agent architectures, including communication, planning, scheduling, execution monitoring, coordination, diagnosis, and learning. From a user/programmer perspective, DECAF distinguishes itself by removing the focus from the underlying components of agent building such as socket creation, message formatting, and agent communication. Instead, users may quickly prototype agent systems by focusing on the domain-specific parts of the problem via a graphical plan editor, reusable generic behaviors [9], and various supporting middle-agents [10]. This paper will briefly describe the key portions of the DECAF toolkit and as well as some of the internal details of the agent execution framework. While not all of the modules have yet been completely realized, DECAF has already been used for teaching purposes, allowing student teams, initially untutored in agent systems, to quickly build prototype multi-agent information gathering systems.
John R. Graham, Keith S. Decker

Section V: Decision Making in a Social Context

Variable Sociability in Agent-Based Decision Making
Abstract
Multi-agent system research is concerned with the issues surrounding the performance of collections of interacting agents. A major concern, therefore, is with the design of the decision making mechanism that the individual agents employ in order to determine which actions to take to achieve their goals. An attractive and much sought after property of this mechanism is that it produces decisions that are rational from the perspective of the individual agent. However, agents are also inherently social. Moreover, individual and social concerns often conflict, perhaps leading to inefficient performance of the individual and the system. To address these problems we propose a formal decision making framework, based on social welfare functions, that combines social and individual perspectives in a unified and flexible manner. The framework is realised in an exemplar computational setting and an empirical analysis is made of the relative performance of varyingly sociable decision making functions in a range of environments.
Lisa Hogg, Nicholas R. Jennings
Cooperation and Group Utility
Abstract
In this paper, we propose an definition of cooperation to shared plans that takes into account the benefit of the whole group, where the group’s benefit is computed by considering also the consequences of an agent’s choice in terms of the actions that the other members of the group will do. In addition, the members of a group consider whether to adopt the goals of their partners: an agent should adopt these goals only when the adoption results in an increase of the group’s benefit.
Guido Boella, Rossana Damiano, Leonardo Lesmo
Relating Quantified Motivations for Organizationally Situated Agents
Abstract
To scale agent technologies for widespread use in open systems, agents must have an understanding of the organizational context in which they operate. In this paper we focus on the issue of task valuation and action selection in socially situated or organized agents – specifically on the issue of quantifying agent relationships and relating work motivated by different sources.
Thomas Wagner, Victor Lesser
The Role and the Impact of Preferences on Multiagent Interaction
Abstract
In the paper we extend our previous analysis of the implications and conditions for employing prosocial and antisocial agents in multiagent interaction. We consider social preferences to be individual preferences that take into account other agents’ preferences. Social preferences can be pro-social and antisocial. We analyze altruistic and envious agents as two generic types of pro-social and antisocial agents. We show that declaring pro- social preferences is not sufficient for obtaining desirable social outcomes. Since every particular situation includes different agents with different preferences, a conflict of preferences could occur even if all agents are pro-social. We provide sufficient conditions for preference consistency in multiagent plans. By consistent preferences we mean preferences that are consistent between agents and consistent for every agent. A negotiation game between different types of agents is analyzed. It is shown that different preferences give rise to different types of strategic behavior.
Sviatoslav Brainov
Deliberative Normative Agents: Principles and Architecture
Abstract
In this paper norms are assumed to be useful in agent societies. It is claimed that not only following norms, but also the possibility of ‘intelligent’ norm violation can be useful. Principles for agents that are able to behave deliberatively on the basis of explicitly represented norms are identified and an architecture is introduced. Using this agent architecture, norms can be communicated, adopted and used as meta-goals on the agent’s own processes. As such they have impact on deliberation about goal generation, goal selection, plan generation and plan selection.
Cristiano Castelfranchi, Frank Dignum, Catholijn M. Jonker, Jan Treur
Backmatter
Metadaten
Titel
Intelligent Agents VI. Agent Theories, Architectures, and Languages
herausgegeben von
Nicholas R. Jennings
Yves Lespérance
Copyright-Jahr
2000
Verlag
Springer Berlin Heidelberg
Electronic ISBN
978-3-540-46467-9
Print ISBN
978-3-540-67200-5
DOI
https://doi.org/10.1007/10719619

Premium Partner

    Bildnachweise