Ebook: Computational Models of Argument

The topic of argumentation, already studied in Antique philosophy, has seen major innovations since the advent of the computer age. Software exists for the creation and evaluation of arguments in high-stake situations, such as medical diagnosis and crime investigation; formal systems help appreciate the role of value judgments underlying opposing positions; and one can enter in argumentative dialogues as if playing a computer game.

Since its start in 2006, the biennial conference series on Computational Models of Argument (COMMA) has been a successful forum for researchers studying argumentation using formal and computational tools. In September 2006, the University of Liverpool organised the first edition. In May 2008, the second was held in France, hosted by the Institut de Recherche en Informatique de Toulouse (IRIT). The third edition was organised by the University of Brescia, and held in Desenzano del Garda, Italy, in September 2010. In 2012, the fourth edition of COMMA is held from September 10–12 in Vienna, Austria.

Argumentation can be studied from many angles. One can aim for the building of smart software (the artificial systems perspective), or for a better understanding of the intricacies of human argument (the natural systems perspective), or for the development of an elegant mathematical model of argument (the theoretical systems perspective). Progress in argumentation research is driven by the cross-fertilization and gradual integration of achievements in each of the perspectives (Figure 1).

These perspectives, and more, are present at the conference. The invited speakers at the conference are representatives of this diversity: Trevor Bench-Capon, a philosopher turned computer scientist studying legal applications; Erik Krabbe, who connects two millenia of insights about argument and dialogue, both informal and formal; and Keith Stenning, an experimental psychologist inspired by nonmonotonic logic and artificial intelligence.

The success of the field is illustrated by the increasing number of submissions: in 2006, around 50; in 2008, 60; in 2010, 67; this year, 76. In order to stimulate interaction between researchers with theoretical and practical research aims, in this fourth edition of COMMA, papers could be submitted both for the regular track and for the innovative applications track, the latter new in this edition. We received 65 regular track papers and 11 innovative applications track papers. 28 of them were accepted as full papers, and 17 as short papers. To further emphasise the importance of implemented systems, we also called for system demonstrations; 13 were accepted for the conference, 3 of them associated with another paper in the proceedings, and 10 described in an extended abstract.

The selection of papers and demonstrations was made on the basis of the scholarly reviews and discussion by the members of the Program Committee and additional reviewers. We thank them all for their hard work. Special thanks go to Adam Wyner for his excellent work as demonstrations coordinator. Finally, we are particularly grateful to all people who helped us in organizing COMMA 2012, in particular Eva Nedoma, Markus Pichlmair, and Friedrich Slivovsky.

We gratefully acknowledge financial support from the Vienna Center for Logic and Algorithms (VCLA), the Wolfgang Pauli Institute (WPI), the Institute of Artificial Intelligence (University of Groningen), the European Network for Social Intelligence (SINTELNET)the COST Action on Agreement Technologies and the Taylor & Francis Group.

Groningen/Vienna, July 2012

Bart Verheij (Program Chair)

Stefan Szeider (Conference Chair)

Stefan Woltran (Conference Chair)

In this paper I review my engagement with argumentation over the past forty years. I describe the perspective I brought from philosophy and the Civil Service, and consider a number of aspects of computational argumentation: knowledge based systems, explanation, context, audiences, schemes and models. A key feature of argumentation is that it is an activity which has to be actively engaged with, whereas a proof is an object to be understood and admired.

In this paper, we present some foundational elements related to argument extraction in opinion texts with the objective to further analyse and synthetise user preferences and value systems as they emerge in such texts. We show that (1) within the context of opinionated expressions, a number of evaluative expressions with a ‘heavy’ semantic load receive an argumentative interpretation and (2) that the association of an evaluative expression with a discourse structure such as an elaboration, an illustration, or a reformulation must also be interpreted as an argument. We develop a conceptual semantics of these relations and show how they are analyzed using the Dislog programming language on the <TextCoop> platform, dedicated to discourse analysis.

The goal of this research is to develop technology to help students learn how to argue cogently on scientific questions. We are currently developing the Genetics Argumentation Inquiry Learning (GAIL) system, which will engage in two-party argumentation on cases in clinical genetics. Use of a computational model of argumentation will enable GAIL to analyze, evaluate, and challenge learner's arguments.

Argumentation is key to understanding and evaluating many texts. The arguments in the texts must be identified; using current tools, this requires substantial work from human analysts. With a rule-based tool for semi-automatic text analysis support, we facilitate argument identification. The tool highlights potential argumentative sections of a text according to terms indicative of arguments (e.g. ‘suppose’ or ‘therefore’) and domain terminology (e.g. camera names and properties). The information can be used by an analyst to instantiate argumentation schemes and build arguments for and against a proposal. The resulting argumentation framework can then be passed to argument evaluation tools.

In this paper we present the EcoBioCap project and the modelling needs of this project in terms of argumentation based preference aggregation. The aim of the paper is to well describe the problem encountered in this context and to propose a preference logic in line with the expressivity needed by the application. We then show how to embed this logic within the ASPIC+ system. Finally, we show how argument by expert opinion could be integrated within our framework where preference aggregation needs to take into consideration the different expertise of the project stakeholders.

The field of cancer research is now generating vast amounts of data from a variety of high throughput techniques and these have helped to define cancers based on their genetic foundations. As this knowledge on the processes and underlying genetics of cancer improve, these should be factored back into the research and analyses conducted by other researchers. Managing this volume of data, often conflicting, is becoming increasingly challenging for researchers. This work demonstrates an innovative application of argumentation theory within cancer research by providing a framework to accommodate missing data, address critical questions and generate hypotheses. The prototype system has been validated to demonstrate it identifies the same interesting interactions and molecules as researchers, even when certain key data was deliberately withheld from the system.

This paper describes work on the use of argumentation tools and techniques to support the policy-deliberation process. The paper first surveys existing research in this area. As will be shown, most research focusses on either the microargumentation or the macro-argumentation aspects of policy-deliberation. However, we claim here that ideally argumentation approaches in policy-deliberation settings should account for both levels together. To this end, this paper introduces a prototype Web-based tool for making sense of both macro-argumentation and micro-argumentation in policy-deliberation, particularly through the use of sophisticated argument visualisation techniques. The paper also describes initial steps towards an RDF specification that combines both macro- and micro-argumentation, which will allow the Web tool to act as an RDF repository for policy-deliberation arguments, with links to the rest of the emerging Linked-Data Web.

The significance of recommender systems has steadily grown in recent years as they help users to access relevant items from the vast universe of possibilities available these days. However, most of the research in recommenders is based purely on quantitative aspects, i.e., measures of similarity between items or users. In this paper we introduce a novel hybrid approach to refine recommendations achieved by quantitative methods with a qualitative approach based on argumentation, where suggestions are given after considering several arguments in favor or against the recommendations. In order to accomplish this, we use Defeasible Logic Programming (DeLP) as the underlying formalism for obtaining recommendations. This approach has a number of advantages over other existing recommendation techniques. In particular, recommendations can be refined at any time by adding new polished rules, and explanations may be provided supporting each recommendation in a way that can be easily understood by the user, by means of the computed arguments.

In this paper, we present a new framework to analyze firewall policy by using argumentation. At the core of this new idea is extending firewall rules with the concept of “reasons” and arguing about the reasons, not the rules. Depending on how the reasons are designed, the resulting framework can be useful in a number of ways: new anomalies in a firewall policy can be identified while, at the same time, stronger recommendations can be given to resolve those anomalies that are detected.

We introduce a unified logical approach, based on signed theories and Quantified Boolean Formulas (QBFs), that can serve as a basis for representing and reasoning with various argumentation-based decision problems. By this, we are able to represent, in a uniform and simple way, a wide range of extension-based semantics for argumentation theory, including complete, grounded, preferred, semi-stable, stage, ideal and eager semantics. Furthermore, our approach involves only propositional languages and quantifications over propositional variables, making decision problems like skeptical and credulous acceptance of arguments simply a matter of logical entailment and satisfiability, which can be verified by existing QBF-solvers.

Recently, there has been a proposal by Dung and Thang and by Li et al to extend abstract argumentation to take uncertainty of arguments into account by assigning a probability value to each argument, and then use this assignment to determine the probability that a set of arguments is an extension. In this paper, we explore some of the assumptions behind the definitions, and some of the resulting properties, of the proposal for probabilistic argument graphs.

Argument aggregation is the problem of combining argumentation frameworks. An argument aggregation procedure takes as input an argument framework for each agent in a system, intuitively representing the beliefs of that agent with respect to a disputed domain of discourse; the output is an argumentation framework that represents the social position on the domain of discourse. There are clear analogies between argument aggregation and the well-known preference aggregation problem, which has been extensively studied in the social choice community. The first contribution of this paper is to apply some of the methodology developed in social choice theory to argument aggregation. After recalling the basic framework of Dung's abstract argument systems, and introducing the argument aggregation problem, we motivate and formally define a collection of axioms that specific argument aggregation procedures might or might not satisfy. The second contribution of the paper is to consider the analysis of argument aggregation procedures with respect to these various axioms. We consider a natural representation for argument aggregation procedures, based on Boolean circuits. We then investigate the problem of verifying whether an argument aggregation procedure, presented in this way, does or does not satisfy a number of the axioms we introduced.

We introduce a family of new equational semantics for argumentation networks which can handle odd and even loops in a uniform manner. We offer equational semantics which is equivalent to CF2 semantics, and a better version which gives the same results as traditional Dung semantics for even loops but can still handle odd loops.

In this work, we develop a game-theoretic framework in which pro and con arguments are put forward. This framework intends to capture winning strategies for different defense criteria. In turn, each possible extension semantics satisfies a particular defense criterion. To ensure that only semantics satisfying given criteria are obtained, protocols for playing have to be defined, being the ensuing winning strategies full characterizations of the corresponding criteria. Admissibility and strong admissibility are two of those criteria proposed in the literature for the evaluation of extension semantics; here, we also introduce two weaker criteria, pairwise and weak cogency, for the evaluation of non-admissible semantics, and we define specific game protocolscapturing them.

We introduce a derivative of Dung's seminal abstract argumentation frameworks (afs) through which distinctive features both of Dung's semantics and so-called “value-based” argumentation frameworks (vafs) may be captured. These frameworks, which we describe as uniform afs, thereby recognise that, in some circumstances, arguments may be deemed acceptable, not only as a consequence of subjective viewpoints (as are modelled by the concept of audience in vafs) but also as a consequence of “value independent” acceptance of other arguments: for example in the case of factual statements. We analyse divers acceptability conditions for arguments in uniform afs and obtain a complete picture for the computational complexity of the associated decision questions. Amongst other results it is shown that reasoning in uniform afs may pose significantly greater computational challenges than either standard or value-based questions, a number of problems being complete for the third level of the polynomial hierarchy.

In this paper we consider a debate game between two players in which a player may provide false or inaccurate arguments as a tactic to win the game. We formulate a debate game using a formal argumentation framework and investigate situation where a player may provide dishonest arguments in the game. We also argue how a player can detect dishonest arguments of the opponent player.

In this work we introduce an extension of an structured argumentation system in which presumptions are fully integrated in the reasoning mechanism. A presumption is defined as a piece of information that is tentatively taken to be true, usually in the absence of acceptable reasons to the contrary. In Presumptive Defeasible Logic Programming (PreDeLP), arguments can be built from both facts and presumptions. We analyze several criteria to compare arguments in PreDeLP, and highlight one possibility that captures the semantic difference between presumptions and facts or defeasible rules. On top of PreDeLP we develop a defeasible approach to argumentation-based diagnostic reasoning. The use of presumptions permits to extend the reasoning beyond factual information, creating in this way a universe of possible scenarios that support or interfere with a claim. We characterize the set of possible scenarios for and against a claim and identify those that best explain the warrant status of the query.

The objective of this paper is to try to fill the gap between: argumentation, electronic institutions and protocols by using a combination of automated synthesis and model checking methods. More precisely, this paper proposes a means of moving rapidly from argument specification to protocol implementation, using an extension of the Argument Interchange Format as the specification language and the Lightweight Coordination Calculus as an implementation language.

For Dung's theory of abstract argumentation, algorithms have been introduced for enumerating all preferred extensions. Two specific approaches have been set out that are based on labeling arguments as: IN, OUT or UNDEC. The purpose of this paper is to improve the two existing approaches by introducing two enhancements. Firstly, we employ two more informative labels. Secondly, by using these additional labels, we describe a new scheme for how the arguments' labels change in the course of computing the preferred extensions. Supported by empirical evaluation, we argue that these modifications accelerate computations. Moreover, we show how to apply the new algorithm in the context of value-based frameworks for persuasive argument, and hence, it appears that the new algorithm is usable in other formalisms extending Dung's model.