Search

This paper presents an argumentation mechanism for reconciling conflicts between planning agents related to plan proposals, which are caused by inconsistencies between basic beliefs regarding the state of the world or the specification of the planning operators. We introduce simple and efficient argument moves that enable discussion about planning steps, and show how these can be integrated into an existing protocol for belief argumentation. The resulting protocol is provably sound with regard to the defeasible semantics of the resulting agreements. We show how argument generation can be treated, for the specific task of argumentation about plans, by replacing the burden of finding proofs in a knowledge base by guided search. [ABSTRACT FROM AUTHOR]

Interest-based negotiation (IBN) is a form of negotiation in which agents exchange information about their underlying goals, with a view to improving the likelihood and quality of a deal. While this intuition has been stated informally in much previous literature, there is no formal analysis of the types of deals that can be reached through IBN and how they differ from those reachable using (classical) alternating offer bargaining. This paper bridges this gap by providing a formal framework for analysing the outcomes of IBN dialogues, and begins by analysing a specific IBN protocol. [ABSTRACT FROM AUTHOR]

Modeling different types of dialog between autonomous agents is becoming an important research issue. Several proposals exist with a clear definition of the dialog protocol, which is the set of rules governing the high level behavior of the dialog. However, things seem different with the notion of strategy. There is no consensus on the definition of a strategy and on the parameters necessary for its definition. Consequently, there are no methodology and no formal models for strategies. This paper argues that a strategy is a decision problem that consists of: i) selecting the type of act to utter at a given step of a dialog, and ii) selecting the content that will accompany the act. The first kind of decision amounts to selecting among all the acts allowed by the protocol, the best option which according to some strategic beliefs of the agent will at least satisfy the most important strategic goals of the agent. The second kind of decision consists of selecting among different alternatives (eg. different offers), the best one that, according to some basic beliefs of the agent, will satisfy the functional goals of the agent. The paper proposes then a formal model based on argumentation for computing on the basis of the above kinds of mental states, the best move (act + content) to play at a given step of the dialog. The model is illustrated through an example of auctions. [ABSTRACT FROM AUTHOR]

This paper presents a coherentist approach to argumentation that extends previous proposals on cognitive coherence based agent communication pragmatics (inspired from social psychology) and propose (1) an alternative view on argumentation that is (2) part of a more general model of communication. In this approach, the cognitive aspects associated to both the production, the evaluation and the integration of arguments are driven by calculus on a formal characterization of cognitive coherence. [ABSTRACT FROM AUTHOR]

In situations where self-interested agents interact repeatedly, it is important that they are endowed with negotiation techniques that enable them to reach agreements that are profitable in the long run. To this end, we devise a novel negotiation algorithm that generates promises of rewards in future interactions, as a means of permitting agents to reach better agreements, in a shorter time, in the present encounter. Moreover, we thus develop a specific negotiation tactic based on this reward generation algorithm and show that it can achieve significantly better outcomes than existing benchmark tactics that do not use such inducements. Specifically, we show, via empirical evaluation, that our tactic can lead to a 26% improvement in the utility of deals that are made and that 21 times fewer messages need to be exchanged in order to achieve this. [ABSTRACT FROM AUTHOR]

While researchers have looked at many aspects of argumentation, an area often neglected is that of argumentation strategies. That is, given multiple possible arguments that an agent can put forth, which should be selected in what circumstances. In this paper, we propose two related heuristics that allow an agent to select what utterances to make. The first involves minimising the amount of information revealed in the course of a dialogue. The second heuristic assigns a utility cost to revealing information, as well as a utility to winning, drawing and losing an argument. An agent participating in a dialogue then attempts to maximise its utility. We present a formal argumentation framework in which these heuristics may operate, and show how they function within the framework. Finally, we discuss extensions to the heuristics, and their relevance to argumentation theory in general. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new strategic and tactic reasoning for agent communication. This reasoning framework is specified using argumentation theory combined to a relevance theory. Strategic reasoning enables agents to decide about the global communication plan in terms of the macro-actions to perform in order to achieve the main conversational goal. Tactic reasoning, on the other hand, allows agents to locally select, at each moment, the most appropriate argument according to the adopted strategy. Previous efforts at defining and formalizing strategies for argumentative agents have often neglected the tactic level and the relation between strategic and tactic levels. In this paper, we propose a formal framework for strategic and tactic reasoning for rational communicating agents and the relation between these two kinds of reasoning. Furthermore, we address the computational complexity of this framework and we argue that this complexity is in the same level of the polynomial hierarchy than the complexity of the strategic-free argumentation reasoning. [ABSTRACT FROM AUTHOR]

Social influences play an important part in the actions that an individual agent may perform within a multi-agent society. However, the incomplete knowledge and the diverse and conflicting influences present within such societies, may stop an agent from abiding by all its social influences. This may, in turn, lead to conflicts that the agents need to identify, manage, and resolve in order for the society to behave in a coherent manner. To this end, we present an empirical study of an argumentation-based negotiation (ABN) approach that allows the agents to detect such conflicts, and then manage and resolve them through the use of argumentative dialogues. To test our theory, we map our ABN model to a multi-agent task allocation scenario. Our results show that using an argumentation approach allows agents to both efficiently and effectively manage their social influences even under high degrees of incompleteness. Finally, we show that allowing agents to argue and resolve such conflicts early in the negotiation encounter increases their efficiency in managing social influences. [ABSTRACT FROM AUTHOR]

Distributed Constraint Satisfaction Problems provide a natural mechanism for multiagent coordination and agreement. To date, algorithms for Distributed Constraint Satisfaction Problems have tended to mirror existing non-distributed global-search or local-search algorithms. Unfortunately, existing distributed global-search algorithms derive from classical backtracking search methods and require a total ordering over agents for completeness. Distributed variants of local-search algorithms (such as distributed breakout) inherit the incompleteness properties of their predecessors, or depend on the creation of new communication links between agents. This paper presents a new approach, inspired by argumentation, to solve DisCSP instances while avoiding some of the identified drawbacks of global- and local-search. [ABSTRACT FROM AUTHOR]

We build on recent work on argumentation frameworks for generating desires and plans. We provide a rich instantiation of Dung's abstract argumentation framework for (i) generating consistent desires; and (ii) generating consistent plans for achieving these desires. This is done through three distinct argumentation frameworks: one (now standard) for arguing about beliefs, one for arguing about what desires the agent should adopt, and one for arguing about what plans to intend in order to achieve the agent's desires. More specifically, we refine and extend existing approaches by providing means for comparing arguments based on decision-theoretic notions (cf. utility). Thus, the worth of desires and the cost of resources are integrated into the argumentation frameworks and taken into account when comparing arguments. [ABSTRACT FROM AUTHOR]

One of the most widely studied systems of argumentation is the one described by Dung in a paper from 1995. Unfortunately, this framework does not allow for joint attacks on arguments, which we argue must be required of any truly abstract argumentation framework. A few frameworks can be said to allow for such interactions among arguments, but for various reasons we believe that these are inadequate for modelling argumentation systems with joint attacks. In this paper we propose a generalization of the framework of Dung, which allows for sets of arguments to attack other arguments. We extend the semantics associated with the original framework to this generalization, and prove that all results in the paper by Dung have an equivalent in this more abstract framework. [ABSTRACT FROM AUTHOR]

Multiagent learning can be seen as applying ML techniques to the core issues of multiagent systems, like communication, coordination, and competition. In this paper, we address the issue of learning from communication among agents circumscribed to a scenario with two agents that (1) work in the same domain using a shared ontology, (2) are capable of learning from examples, and (3) communicate using an argumentative framework. We will present a two fold approach consisting of (1) an argumentation framework for learning agents, and (2) an individual policy for agents to generate arguments and counterarguments (including counterexamples). We focus on argumentation between two agents, presenting (1) an interaction protocol (AMAL2) that allows agents to learn from counterexamples and (2) a preference relation to determine the joint outcome when individual predictions are in contradiction. We present several experiment to asses how joint predictions based on argumentation improve over individual prediction. [ABSTRACT FROM AUTHOR]

Computational argumentation has been accepted as a social computing mechanism or paradigm in the multi-agent systems community. In this paper, we are further concerned with what agents believe after argumentation, such as how agents should accommodate justified arguments into their knowledge bases after argumentation, what and how agents acquire or learn, based on the results of argumentation. This is an attempt to create a new learning method induced by argumentation that we call Argument-Based Learning (ABL). To this end, we use our logic of multiple-valued argumentation LMA built on top of Extended Annotated Logic Programming EALP, and propose three basic definitions to capture agents' beliefs that should be rationally acquired after argumentation: knowledge acquisition induced by the undercut of assumptions, knowledge acquisition induced by difference of recognition, and knowledge acquisition induced by rebut. They are derived from two distinctive and advantageous apparatuses of our approach to multi-valued argumentation under : Paraconsistency and multiple-valuedness that EALP and LMA feature. We describe an overall argument example to show the effectiveness and usefulness of the agent learning methods based on argumentation. [ABSTRACT FROM AUTHOR]

This chapter provides a brief survey of argumentation in multi-agent systems. It is not only brief, but rather idiosyncratic, and focuses on the areas of research that most interest the authors, and those which seem to be the most active at the time of writing. [ABSTRACT FROM AUTHOR]

In our previous work on dialogue games for agent interaction, an agent's set of beliefs (Σ) and an agent's "commitment store" (CS) — the set of locutions uttered by the agent — play a crucial role. The usual assumption made in this work is that the set of beliefs is static through the course of a dialogue, while the commitment store is dynamic. While the assumption of static beliefs is reasonable during the progress of the dialogue, it seems clear that some form of belief change is appropriate once a dialogue is complete. What form this change should take is our subject in this paper. [ABSTRACT FROM AUTHOR]

Disseminating pieces of knowledge among the members of large organizations is a well known problem in Knowledge Management, involving several decision-making processes. The JITIK multiagent framework has been successfully used for just-in-time delivering highly customized notifications to the adequate users in large distributed organizations. However, in JITIK as well as in other similar approaches it is common to deal with incomplete information and conflicting policies, making difficult to make decisions about whether to deliver or not a specific piece of information or knowledge on the basis of a rationally justified procedure. This paper presents an approach to cope with this problem by integrating JITIK with a defeasible argumentation formalism. Conflicts among policies are solved on the basis of a dialectical analysis whose outcome determines whether a particular information item should be delivered to a specific user. Keywords: Argumentation, knowledge management, information systems. [ABSTRACT FROM AUTHOR]

This paper aims at proposing a general formal framework for dialogue between autonomous agents which are looking for a common agreement about a collective choice. The proposed setting has three main components: the agents, their reasoning capabilities, and a protocol. The agents are supposed to maintain beliefs about the environment and the other agents, together with their own goals. The beliefs are more or less certain and the goals may not have equal priority. These agents are supposed to be able to make decisions, to revise their beliefs and to support their points of view by arguments. A general protocol is also proposed. It governs the high-level behaviour of interacting agents. Particularly, it specifies the legal moves in the dialogue. Properties of the framework are studied. This setting is illustrated on an example involving three agents discussing the place and date of their next meeting. Keywords: Argumentation, Negotiation. [ABSTRACT FROM AUTHOR]

A key goal for a scientist is to find evidence to argue for or against universal statements (in effect first-order formulae) about the world. Building logic-based tools to support this activity could be potentially very useful for scientists to analyse new scientific findings using experimental results and established scientific knowledge. In effect, these logical tools would help scientists to present arguments and counterarguments for tentative scientific knowledge, and to share and discuss these with other scientists. To address this, in this paper, we explain how tentative and established scientific knowledge can be represented in logic, we show how first-order argumentation can be used for analysing scientific knowledge, and we extend our framework for evaluating the degree of conflict arising in scientific knowledge. We also discuss the applicability of recent developments in optimizing the impact and believability of arguments for the intended audience. [ABSTRACT FROM AUTHOR]

This paper presents an argumentation-based approach to deliberation, the process by which two or more agents reach a consensus on a course of action. The kind of deliberation that we are interested in is a process that combines both the selection of an overall goal, the reduction of this goal into sub-goals, and the formation of a plan to achieve the overall goal. We develop a mechanism for doing this, describe how this mechanism can be integrated into a system of argumentation to provide a sound and complete deliberation system, and show how the same process can be achieved through a multi-agent dialogue. [ABSTRACT FROM AUTHOR]

Autonomous agents working in multi-agent environments need to cooperate in order to fulfill tasks. Generally, an agent cannot perform a task alone and needs help from the other agents. One of the solutions to this problem is to look for groups of agents which are able to perform the desired tasks better. Different algorithms have then been proposed for the task allocation via coalition formation. This last is generally seen as a two steps process: i) constructing the different solutions (called coalitions structures), then ii) discussing these solutions between the agents in order to select the best ones which will be adopted. This paper provides a unified formal framework for constructing the coalitions structures. In fact, we will show that the problem of coalition formation can be defined in terms only of a set of coalitions whose structures are abstract, a conflict relationship between the coalitions and a preference relation between the coalitions. Three semantics for coalitions structures will be proposed: a basic semantics which will return a unique coalition structure, stable semantics and preferred semantics. These two last may return several coalitions structures at the same time. A proof theory of the basic semantics will also be proposed. The aim of this proof theory is to test whether a given coalition will be acceptable for the agent or not without computing the whole structure. We will show that this framework is general enough to capture different propositions made in the literature. An instantiation of our framework is given and its properties are studied. [ABSTRACT FROM AUTHOR]

Inspired by computational linguistic approaches to annotate the structures that occur in human dialogue, this paper describes a technique which encodes these structures as transformations applied to a protocol language. Agents can have a controlled and verifiable mechanism to synthesise and communicate their interaction protocol during their participation in a multiagent system. This is in contrast to the approaches where agents must subscribe to a fixed protocol and relinquish control over an interaction that may not satisfy the agent's dialogical needs or rely on internal its reasoning to determine which message to communicate at a certain point in the dialogue. [ABSTRACT FROM AUTHOR]

This paper proposes a liberalized version of existing truth-finding protocols for argumentation, such as the standard two-agent immediate-response protocol for computing the credulous acceptance of conclusions in an argument system. In the new setup agents decide autonomously which issues need to be discussed, when to query other agents, when to keep on querying other agents, and when to settle for an answer. In this way, inter-agent disputes are regulated by the agents themselves, rather than by following an outlined protocol. The paper concludes with a prototype implementation and with a comparison of related work on conversation analysis and computational dialectic.A(colorful and instructive) poster based on a shorter version of this paper was presented at AAMAS'05 . The poster itself can be viewed at http://www.cs.uu.nl/~gv/abstracts/liberal_protocol_poster.pdf. [ABSTRACT FROM AUTHOR]

This paper presents a novel approach to modelling embedded agent dialogues. It proposes a specific structure for the supporting information accompanying the arguments that agents exchange during a dialogue, it defines formally how this information relates to the agent theory, and assigns to it semantics that is associated to each of the atomic dialogue types of the Walton-Krabbe typology. This allows the formal definition of necessary and sufficient initiation and acceptance conditions of licit dialectical shifts that are necessary for the modelling of embedded agent dialogues. [ABSTRACT FROM AUTHOR]

Much work in the area of argumentation-based dialogues between agents has been based on the influential taxonomy of dialogue types developed by Walton and Krabbe. In this paper we re-examine the Walton and Krabbe framework, concentrating on the preconditions for different types of dialogue and analyzing them in a systematic way. Doing so identifies a number of new kinds of dialogue missing from the framework. We discuss some of the more interesting of these and develop protocols for them. [ABSTRACT FROM AUTHOR]

One form of argument-based negotiation is when agents argue about why an offer was rejected. If an agent can state a reason for a rejection of an offer, the negotiation process may become more efficient since the other agent can take this reason into account when making new offers. Also, if a reason for rejection can be disputed, the negotiation process may be of higher quality since flawed reasons may be revised as a result. This paper presents a formal protocol for negotiation in which reasons can be asked and given for rejections and in which agents can try to persuade each other that a reason is or is not acceptable. The protocol is modelled as a persuasion dialogue game embedded in a negotiation protocol. It has a social semantics since the protocol does not refer to the internal state of negotiating agents. [ABSTRACT FROM AUTHOR]

Recent years have seen an increasing interest of multiagent system research in employing the theory of argumentation for the development of communication protocols. While significant progress has been made in formalising argument-based communication, (possibly adaptive) agent-level argumentation strategies as a practical integration of rational agent reasoning and inter-agent argumentation dialogues have received fairly little attention. In this paper we propose the use of the InFFrA framework in argument-based negotiation. This framework allows for a strategic and adaptive communication to achieve private goals within the limits of bounded rationality in open argumentation communities. The feasibility of the approach is illustrated in an agent-based web linkage scenario, showing that its performance is comparable to that of simple proposal-based negotiation while accommodating much stricter constraints regarding "what can be said" like those used in argumentation. [ABSTRACT FROM AUTHOR]

Argumentation-based negotiation (ABN) provides agents with an effective means to resolve conflicts within a multi-agent society. However, to engage in such argumentative encounters the agents require the ability to generate arguments, which, in turn, demands four fundamental capabilities: a schema to reason in a social context, a mechanism to identify a suitable set of arguments, a language and a protocol to exchange these arguments, and a decision making functionality to generate such dialogues. This paper focuses on the first two issues and formulates models to capture them. Specifically, we propose a coherent schema, based on social commitments, to capture social influences emanating from the roles and relationships of a multi-agent society. After explaining how agents can use this schema to reason within a society, we then use it to identify two major ways of exploiting social influence within ABN to resolve conflicts. The first of these allows agents to argue about the validity of each other's social reasoning, whereas the second enables agents to exploit social influences by incorporating them as parameters within their negotiation. For each of these, we use our schema to systematically capture a comprehensive set of social arguments that can be used within a multi-agent society. Keywords: Argumentation-based Negotiation, Conflict Resolution. [ABSTRACT FROM AUTHOR]