Invited Speakers

• Gerhard Brewka, University of Leipzig, Germany.


• Laura Kovacs, TU Wien, Austria.


• Sebastian Link, University of Auckland, New Zealand.


• David Pearce, Technical University of Madrid, Spain (supported by ALP).


• Bernhard Thalheim, Christian-Albrechts-University at Kiel Germany.



Computational Models of Argument: A New Perspective on Persisting KR Problems

Joachim Biskup

Gerhard Brewka, University of Leipzig, Germany.

Gerhard Brewka is a Professor of Intelligent Systems at Leipzig University, Germany. His research focuses on knowledge representation, in particular logic programming, nonmonotonic reasoning, preference and inconsistency handling, and computational models of argumentation. He served as President of EurAI (formerly ECCAI), the European Association of AI, and of Knowledge Representation Inc. In 2002, Brewka was awarded a EurAI Fellowship. He is a member of the IJCAI Board of Trustees and was Conference Chair of IJCAI-16 in New York.

In the last two decades the area of knowledge representation and reasoning (KR) has seen a steady rise of interest in the notion of argument, an old topic of study in philosophy. This interest was fueled by a certain dissatisfaction with existing approaches, especially to default reasoning and inconsistency handling, and by the demands of applications in legal reasoning and several related fields.

The ultimate goal of computational argumentation is to enable the development of computer-based systems capable to support -- and to participate in -- argumentative activities.
To achieve this goal one has to come up with models which formally capture the way we usually come to conclusions and make decisions, namely by

 1.- Constructing arguments for and against various options,
 2.- Establishing relationships among the arguments, most notably the attack relation, and
 3.- Identifying interesting subsets of the arguments which represent coherent positions based on these relations.

In the talk we will highlight some of the main ideas and key techniques that have been developed in the field and show how they address issues of representing knowledge, handling inconsistencies, and reasoning by default. We will mainly focus on Abstract Dialectical Frameworks (ADFs) which substantially generalize the well-known and widely used Dung Frameworks. In particular, we will demonstrate how the operator-based techniques underlying ADFs allow us to turn directed graphs with arbitrary edge labels, which are widely used to visualize argumentation and reasoning scenarios, into full-fledged knowledge representation formalisms with a whole range of precisely defined semantics.

Automated Reasoning for Systems Engineering

Christoph Beierle

Laura Kovacs, TU Wien, Austria.

Laura Kovács is a full professor at the Faculty of Informatics of Vienna University of Technology (TU Vienna). She also holds a part-time professor position at the Department of Computer Science and Engineering of the Chalmers University of Technology. She has a diploma in computer science and math from the West University of Timisoara, Romania and a PhD with highest distinction in computer science from the Research Institute of Symbolic Computation (RISC-Linz) of the Johannes Kepler University Linz, Austria. Prior to her appointment to Vienna, she was an associate professor at Chalmers.

In her research, Laura Kovács deals with the design and development of new theories, technologies, and tools for program analysis, with a particular focus on automated assertion generation, symbolic summation, computer algebra, and automated theorem proving. She is the co-developer of the Vampire theorem prover. In 2014, she received the Wallenberg Academy Fellowship and an ERC Starting Grant.

Automated reasoning, and in particular first-order theorem proving, is one of the earliest research areas within artificial intelligence and formal methods. It is undergoing a rapid development thanks to its successful use in program analysis and verification, semantic Web, database systems, symbolic computation, theorem proving in mathematics, and other related areas. Breakthrough results in all areas of theorem proving have been obtained, including improvements in theory, implementation, and the development of powerful theorem proving tools.

In this talk I give a brief overview on the main ingredients of automated theorem proving, and focus on recent challenges and developments in the area. Further, I will discuss recent applications of theorem proving in rigorous systems engineering.

Old keys that open new doors

Reinhard Pichler

Sebastian Link, University of Auckland, New Zealand.

Sebastian Link received a DSc from the University of Auckland in 2015, and a PhD in Information Systems from Massey University in 2005. Currently, he is a full professor at the Department of Computer Science in the University of Auckland. His research interests include conceptual data modeling, semantics in databases, foundations of mark-up languages, and applications of discrete mathematics to computer science. Sebastian received the Chris Wallace Award for Outstanding Research Contributions from the Computing Research and Education Association of Australasia in recognition of his work on the semantics of SQL and XML data. Sebastian has published more than 150 research articles, and served as a reviewer for numerous conferences and journals. He is a member of the editorial board of the journal Information Systems.

Keys enforce Codd's integrity for entities,
Giving fast access to data since the seventies.
The issue of missing information remains fundamental,
Better notions of keys will prove to be instrumental.

We review keys on classical relations,
Recalling the simplest of all axiomatisations.
An extremal cardinality a non-redundant family retains,
Whenever it lives up to Sperner's anti-chains.
Armstrong relations are built after an anti-key hunt,
The discovery by hypergraph transversals is simply elegant.

As nulls in applications do require some finesse,
We review key sets that have high expressiveness.
Establishing an axiomatisation that is binary,
We show implication to be complete for coNP.
Armstrong relations do not necessarily exist,
The discovery of keys sets as an open problem we enlist.

Key sets with singletons avoid the likely intractability curtain,
Leading to keys that hold in every world so certain.
We look at possible and certain keys together with NOT NULL,
Which lead to problems that are anything but dull.
Implication is easily characterised axiomatically and algorithmically,
The structure and computation of Armstrong relations is captured non-trivially.
Extremal families occupy two levels with some gaps,
The discovery can use transversals in two steps.

We briefly summarise keys on data with veracities,
considering probabilities, possibilities, and contextualities.
Concluding with probems for minds that are bright,
We hope the talk sparks research with heaps of insight.

The Logical Basis of Knowledge Representation in Answer Set Programming

Henry Prakken

David Pearce, Technical University of Madrid, Spain (supported by ALP).

David Pearce studied Philosophy, Logic and Scientific Method at the Universities of Sussex and Oxford, obtaining his D Phil (Sussex) in 1980. From 1982-94 he worked at the Philosophy Institute of the Free University Berlin as a Lecturer and later Heisenberg Research Fellow. From 1992-94 he was Acting Professor at the Universities of Göttingen and Heidelberg. In 1994 he moved to the German AI Research Centre (DFKI) in Saarbrücken, where until 2000 he coordinated one of the founding European Networks of Excellence: Compulog Net. From 2000-2002 he worked at the Future and Emerging Technologies Unit of the European Commission in Brussels where he was involved in the management and supervision of EU research programmes. He then moved to Madrid as Ramón y Cajal Research Fellow at the Rey Juan Carlos University, later becoming professor in the Technical University of Madrid in 2009. From 2011-14 he coordinated the EU funded action: the European Network for Social Intelligence (SINTELNET). David Pearce has worked mainly in the areas of Logic and Knowledge Representation, with a special interest in nonmonotonic reasoning and logic programming. He has made numerous contributions to the field of Answer Set Programming (ASP). In the late 1980s, together with Gerd Wagner, he introduced the concept of strong negation into logic programming. From 1995 onwards he developed Equilibrium Logic as a new logical foundation for ASP. In 2001, together with Vladimir Lifschitz and Agustín Valverde, he initiated the study of strongly equivalent logic programs which opened up a new researcharea in nonmonotonic reasoning and KRR that is still active today. His current  research interests include combining Artificial Intelligence with Social Ontology. Pearce was elected ECCAI (now EurAi) Fellow in 2014.

In this talk I give an introduction to the underlying logic of Answer Set Programming. The basis is a non-classical, intermediate logic and its non-monotonic extension, known as equilibrium logic. Together they provide analternative to the standard paradigm of two-valued, classical logic. In view of the origins of answer set programming, it seems appropriate to call this new approach stable reasoning.

The talk will focus on the intuitive meaning of the main logical definitions, and explain their effect with some example programs. I will also discuss some of the main extensions of the basic language that may be useful for knowledge representation.

Revisiting the Database Constraints Theory

José María Turull Torres

Bernhard Thalheim, Christian-Albrechts-University at Kiel Germany.

Prof. Dr.rer.nat.habil. Bernhard Thalheim (Director, Department of Computer Science, Faculty of Engineering at Christian-Albrechts University Kiel, Germany) (MSc, PhD, DSc) is full professor at Christian Albrechts University in Germany. His major research interests are database theory, logic in databases, and systems development methodologies, in particular for web information systems. He has published more than 300 refereed publications, edited more than 30 conference volumes, co-founded three international conferences, and has been programme committee chair for almost three dozen international conferences such as MFDBS, ER, FoIKS, ASM, SDKB, NLDB and ADBIS. He got several international awards, e.g. the Kolmogorov professorship at Lomonossow University Moscov and the P.P. Chen award of Elsevier. He has been an associated professor at Dresden University of Technology, a visiting professor at Kuwait University, Alpen-Adria University Klagenfurt and others, and a full professor at Rostock University and Brandenburg University of Technology at Cottbus.

The theory of database constraints has been developed for a long time within the relational database modelling setting. The 80ies brought a large body of knowledge and led to the impression that the theory development is completed. A typical example is normalisation theory that has been developed inside the relational understanding. It must already reconsidered for the table database modelling setting. Cardinality constraints defined in an entity-relationship modelling setting were the most essential addition to the theory of relational constraints. It seems that the theory of object-relational constraints is still a lacuna. Therefore, monographs and textbooks remain to be on the level of the early 90ies as far as constraints are considered. Database technology brought however powerful and sophisticated systems. So, the constraints that might be supported without loss of performance are far richer. Database applications need more sophisticated constraints. So, the paper presents some solutions for constraint enhancement, constraint handling, structure optimisation, and database modelling at the conceptual level. It completes with open problems.