nach oben

2012 | Buch

Kapitel lesen Erstes Kapitel lesen

Theorie and Applications of Formal Argumentation

First International Workshop, TAFA 2011. Barcelona, Spain, July 16-17, 2011, Revised Selected Papers

herausgegeben von: Sanjay Modgil, Nir Oren, Francesca Toni

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

Über dieses Buch

This book constitutes the thoroughly refereed proceedings of the First International Workshop on the Theory and Applications of Formal Argumentation, TAFA 2011, held in Barcelona, Spain, in Juli 2011, as a workshop at IJCAI 2011, the 22nd International Joint Conference on Artificial Intelligence. The 9 revised full papers presented together with 8 revised poster papers were carefully selected during two rounds of reviewing and improvement from 32 initial submissions. The workshop promotes and fosters uptake of argumentation as a viable AI paradigm with wide ranging application, and provides a forum for further development of ideas and the initiation of new and innovative collaborations. The papers cover the following topics: properties of formal models of argumentation; instantiations of abstract argumentation frameworks; relationships among different argumentation frameworks; practical applications of formal models of argumentation; argumentation and other artificial intelligence techniques; evaluation of formal models of argumentation; validation and evaluation of applications of argumentation.

Inhaltsverzeichnis

Frontmatter

Theory and Applications of Formal Argumentation

Probabilistic Argumentation Frameworks

Abstract

In this paper, we extend Dung’s seminal argument framework to form a probabilistic argument framework by associating probabilities with arguments and defeats. We then compute the likelihood of some set of arguments appearing within an arbitrary argument framework induced from this probabilistic framework. We show that the complexity of computing this likelihood precisely is exponential in the number of arguments and defeats, and thus describe an approximate approach to computing these likelihoods based on Monte-Carlo simulation. Evaluating the latter approach against the exact approach shows significant computational savings. Our probabilistic argument framework is applicable to a number of real world problems; we show its utility by applying it to the problem of coalition formation.

Hengfei Li, Nir Oren, Timothy J. Norman

Splitting Argumentation Frameworks: An Empirical Evaluation

Abstract

In a recent paper Baumann [1] has shown that splitting results, similar to those known for logic programs under answer set semantics and default logic, can also be obtained for Dung argumentation frameworks (AFs). Under certain conditions a given AF A can be split into subparts A ₁ and A ₂ such that extensions of A can be computed by (1) computing an extension E ₁ of A ₁, (2) modifying A ₂ based on E ₁, and (3) combining E ₁ and an extension E ₂ of the modified variant of A ₂. In this paper we perform a systematic empirical evaluation of the effects of splitting on the computation of extensions. Our study shows that the performance of algorithms may drastically improve when splitting is applied.

Ringo Baumann, Gerhard Brewka, Renata Wong

On the Complexity of Computing the Justification Status of an Argument

Abstract

We address the problem of determining the acceptance status of an argument w.r.t. labeling-based semantics. Wu and Caminada recently proposed a labeling-based justification status of arguments to distinguish different levels of acceptability for arguments. We generalize their approach, which was originally restricted to complete semantics, to arbitrary argumentation semantics and provide a comprehensive study of the computational properties.

Wolfgang Dvořák

Arguments over Co-operative Plans

Abstract

Autonomous planning agents that share a common goal should be able to propose, justify and share information about plans. To reach an agreement on the best plan, strategies for persuasion and negotiation could be used by agents in order to share their beliefs about the world and resolve conflicts between the agents. We present an argumentation scheme and associated critical questions to create and justify plan proposals where plans are combinations of actions requiring several agents for their execution. An analysis of different ways in which actions can combine is presented and then associated with the argumentation scheme and the critical questions. We believe these elements are necessary to enable agents to engage in rational debate over co-operative plan proposals.

Rolando Medellin-Gasque, Katie Atkinson, Peter McBurney, Trevor Bench-Capon

An Implemented Dialogue System for Inquiry and Persuasion

Abstract

In this paper, we present an implemented system that enables autonomous agents to engage in dialogues that involve inquiries embedded within a process of practical reasoning. The implementation builds upon an existing formal model of value-based argumentation, which has itself been extended to permit a wider range of arguments to be expressed. We present extensions to the formal underlying theory used for the dialogue system, as well as the implementation itself. We demonstrate the use of the system through a particular case study. We discuss a number of interesting issues that have arisen from the implementation and the experimental avenues that this test-bed will enable us to pursue.

Luke Riley, Katie Atkinson, Terry Payne, Elizabeth Black

An Argumentation Framework for Qualitative Multi-criteria Preferences

Abstract

Preferences between different alternatives (products, decisions, agreements etc.) are often based on multiple criteria. Qualitative Preference Systems (QPS) is a formal framework for the representation of qualitative multi-criteria preferences in which a criterion’s preference is defined based on the values of attributes or by combining multiple subcriteria in a cardinality-based or lexicographic way. In this paper we present a language and reasoning mechanism to represent and reason about such qualitative multi-criteria preferences. We take an argumentation-based approach and show that the presented argumentation framework correctly models a QPS. Then we extend this argumentation framework in such a way that it can derive missing information from background knowledge, which makes it more flexible in case of incomplete specifications.

Wietske Visser, Koen V. Hindriks, Catholijn M. Jonker

Modeling and Solving AFs with a Constraint-Based Tool: ConArg

Abstract

ConArg is a tool based on Constraint Programming which is able to model and solve different problems related to Argumentation Frameworks (AFs). To practically implement the tool, we have used JaCoP, a Java library which provides the user with a Finite Domain Constraint Programming paradigm. Constraint Satisfaction Problems (CSPs) offer a wide number of efficient techniques (as inference and search algorithms) that can tackle the complexity in finding all the possible Dung’s conflict-free, admissible, complete and stable extensions in AFs. Moreover, we can use the tool to solve some of the preference-based problems presented in literature. ConArg is able to randomly generate networks with small-world properties in order to find Dung’s extensions on such interaction graphs. We present the main features of ConArg and we report the performance in time.

Stefano Bistarelli, Francesco Santini

Resource Boundedness and Argumentation

Abstract

In this paper we extend the traditional Dung argumentation framework with cardinality constraints over the set of warranted arguments. This results in a new definition for argumentation semantics wherein arguments within an extension are both in some sense consistent and compliant with the constraints imposed on the system. After discussing the theoretical aspects of such a resource-bounded argumentation framework we describe its utility via an application to a concrete application domain: the scheduling of demand responsive transport.

Nicolás D. Rotstein, Nir Oren, Timothy J. Norman

An Empirical Study of a Deliberation Dialogue System

Abstract

We present an empirical simulation-based study of the use of value-based argumentation in two-party deliberation dialogues, investigating the impact that argumentation can have on the quality of the outcome reached. Our simulation allows us to vary the number of values, actions and arguments that appear in the system; we investigate how the behaviour of the system changes as these parameters vary. This parameter sensitivity analysis tells us whether a value-based deliberation dialogue system may be useful for a particular real-world application. We measure the quality of the dialogue outcome (i.e. the action that the agents agree to) against a global view of whether that action would be agreeable to each agent if all of the agents’ knowledge were taken into account. We compare the deliberation outcome with a simple consensus forming procedure (where no arguments are exchanged). Our results show that the deliberation dialogue system we present outperforms consensus forming.

Elizabeth Black, Katie Bentley

Selective Revision by Deductive Argumentation

Abstract

The success postulate of classic belief revision theory demands that after revising some beliefs with by information the new information is believed. However, this form of prioritized belief revision is not apt under many circumstances. Research in non-prioritized belief revision investigates forms of belief revision where success is not a desirable property. Herein, selective revision uses a two step approach, first applying a transformation function to decide if and which part of the new information shall be accepted, and second, incorporating the result using a prioritized revision operator. In this paper, we implement a transformation function by employing deductive argumentation to assess the value of new information. Hereby we obtain a non-prioritized revision operator that only accepts new information if believing in the information is justifiable with respect to the beliefs. By making use of previous results on selective revision we prove that our revision operator satisfies several desirable properties. We illustrate the use of the revision operator by means of examples and compare it with related work.

Patrick Krümpelmann, Matthias Thimm, Marcelo A. Falappa, Alejandro J. García, Gabriele Kern-Isberner, Guillermo R. Simari

A Three-Layer Argumentation Framework

Abstract

Argumentation frameworks which are abstract are suitable for the study of independent properties of any specific aspect (e.g. arguments sceptical and credulous admissible) that are relevant for any argumentation context. However, its direct adoption on specific application contexts requires dealing with questions such as the argument structure, the argument categories, the conditions under which an attack/support is established between arguments, etc. This paper presents a generic argumentation framework which comprehends a conceptualization layer to capture the expressivity and semantics of the argumentation data employed in a specific context and simplifies its adoption by applications. The conceptualization layer together with the defined argument structure is exploited to automatically derive the attack and support relationships between arguments.

Paulo Maio, Nuno Silva

Stable Extensions in Timed Argumentation Frameworks

Abstract

A Timed Abstract Argumentation Framework is a novel formalism where arguments are only valid for consideration in a given period of time, which is defined for every individual argument. Thus, the attainability of attacks and defenses is related to time, and the outcome of the framework may vary accordingly. In this work we study the notion of stable extensions applied to timed-arguments. The framework is extended to include intermittent arguments, which are available with some repeated interruptions in time.

Maria Laura Cobo, Diego C. Martinez, Guillermo R. Simari

Computing with Infinite Argumentation Frameworks: The Case of AFRAs

Abstract

In recent years a large corpus of studies has arisen from Dung’s seminal abstract model of argumentation, including several extensions aimed at increasing its expressiveness. Most of these works focus on the case of finite argumentation frameworks, leaving the potential practical applications of infinite frameworks largely unexplored. In the context of a recently proposed extension of Dung’s framework called AFRA (Argumentation Framework with Recursive Attacks), this paper makes a first step to fill this gap. It is shown that, under some reasonable restrictions, infinite frameworks admit a compact finite specification and that, on this basis, computational problems which are tractable for finite frameworks may preserve the same property in the infinite case. In particular we provide a polynomial-time algorithm to compute the finite representation of the (possibly infinite) grounded extension of an AFRA with infinite attacks. An example concerning the representation of a moral dilemma is introduced to illustrate and instantiate the proposal and gives a preliminary idea of its potential applicability.

Pietro Baroni, Federico Cerutti, Paul E. Dunne, Massimiliano Giacomin

Multi-sorted Argumentation

Abstract

In the theory of abstract argumentation, the acceptance status of arguments is normally determined for the complete set of arguments at once, under a single semantics. However, this is not always desired. In this paper, we extend the notion of an argumentation framework to a multi-sorted argumentation framework, and we motivate this extension using an example which considers practical and epistemic arguments. In a multi-sorted argumentation framework, the arguments are partitioned into a number of cells, where each cell is associated with a semantics under which its arguments are evaluated. We prove the properties of the proposed framework, and we demonstrate our theory with a number of examples. Finally, we relate our theory to the theory of modal fibring of argumentation networks.

Tjitze Rienstra, Alan Perotti, Serena Villata, Dov M. Gabbay, Leendert van der Torre

Conditional Labelling for Abstract Argumentation

Abstract

Agents engage in dialogues having as goals to make some arguments acceptable or unacceptable. To do so they may put forward arguments, adding them to the argumentation framework. Argumentation semantics can relate a change in the framework to the resulting extensions but it is not clear, given an argumentation framework and a desired acceptance state for a given set of arguments, which further arguments should be added in order to achieve those justification statuses. Our methodology, called conditional labelling, is based on argument labelling and assigns to each argument three propositional formulae. These formulae describe which arguments should be attacked by the agent in order to get a particular argument in, out, or undecided, respectively. Given a conditional labelling, the agents have a full knowledge about the consequences of the attacks they may raise on the acceptability of each argument without having to recompute the overall labelling of the framework for each possible set of attack they may raise.

Guido Boella, Dov M. Gabbay, Alan Perotti, Leendert van der Torre, Serena Villata

Bottom-Up Argumentation

Abstract

Online social platforms, e-commerce sites and technical fora support the unfolding of informal exchanges, e.g. debates or discussions, that may be topic-driven or serendipitous. We outline a methodology for analysing these exchanges in computational argumentation terms, thus allowing a formal assessment of the dialectical validity of the positions debated in or emerging from the exchanges. Our methodology allows users to be engaged in this formal analysis and the assessment, within a dynamic process where comments, opinions, objections, as well as links connecting them, can all be contributed by users.

Francesca Toni, Paolo Torroni

A First Step towards Argumentation Dialogues for Discovery

Abstract

We present a formal model for two-agent discovery dialogues. The model allows agents to collectively discover a realization for a shared goal, using argumentation dialogues to exchange information. This information is in the form of rules, assumptions, and contraries of assumptions as in Assumption-based Argumentation (ABA). With dialogues, agents jointly build arguments and construct shared ABA frameworks. We define successful discovery dialogues as those giving admissible arguments that realize the shared goal. The main novelty of this paper is the modelling of the buttom-up relation between utterances. This new relation helps building “higher level” arguments from existing “lower level” supports, which we deem essential for discovery.

Xiuyi Fan, Francesca Toni

Backmatter

Titel: Theorie and Applications of Formal Argumentation
herausgegeben von: Sanjay Modgil
Nir Oren
Francesca Toni
Verlag: Springer Berlin Heidelberg
Electronic ISBN: 978-3-642-29184-5
Print ISBN: 978-3-642-29183-8
DOI: https://doi.org/10.1007/978-3-642-29184-5