Program (preliminary)

Click here for accessing the proceedings of the Conference.

October, 4

08:30-09:30 - Registration

09:30-09:45 - Opening

09:45-10:45 - Keynote 1: Fabio Cozman

10:45-11:15 - Coffee break

11:15-13:15 - Session 1

13:15-14:30 - Lunch

14:30-16:00 - Tutorials 1

16:00-16:30 - Coffee break

16:30-18:30 - Session 2

20:00-22:00 - Welcome cocktail

October, 5

09:00-10:00 - Keynote 2: Alberto Bugarín

10:00-10:45 - Tutorials 2

10:45-11:15 - Coffee break

11:15-13:15 - Session 3

13:15-14:30 - Lunch

14:30-16:00 - Tutorials 3

16:00-20:30 - Cultural Time

20:30-23:30 - Gala Dinner

October, 6

09:00-10:00 - Keynote 3: Martin Theobald

10:00-10:30 - Coffee break

10:30-11:50 - Session 4

11:50-13:20 - Tutorials 4

13:20-14:30 - Lunch

14:30-15:50 - Session 5

15:50-16:35 - Tutorials 5

16:35-17:00 - Coffee break

17:00-18:20 - Session 6

18:20-18:30 - Closing

Keynote 1 (October, 4 - 09:45-10:45)

Chair: Serafín Moral

Fabio Cozman, University of Sao Paulo, Brazil

Large-Scale Relational Bayesian Networks: Specification, Analysis, and Synthesis

There are now several languages that one can use to specify Bayesian networks with repetitive sub-structures, so as to encode large-scale problems such as social networks and knowledge bases.These languages typically resort to predicates and quantifiers, and often to recursive operators. Here we examine such languages, trying to extract their core features, and discuss how we can analyze them, and how we can synthesize (that is, elicit or learn) models using them. We study the connection between logical expressivity and computational complexity, and the challenges we face when learning models. We then consider languages based on logic programming; we again look at their expressivity and complexity. Languages based on logic programming actually go beyond Bayesian networks, so we consider the kinds of probabilistic models one can get with them.


Session 1 (October, 4 - 11:15-13:15)

Chair: Juan M. Fernández-Luna

(11:15) Leopoldo Bertossi

The Causality/Repair Connection in Databases: Causality-Programs

(11:35) Nicolás Marín, Gustavo Rivas-Gervilla, and Daniel Sánchez

Using k-specificity for the management of count restrictions in flexible querying

(11:55) Luis M. de Campos, Juan M. Fernández-Luna, Juan F. Huete and Luis Redondo-Expósito

Comparing Machine Learning and Information Retrieval-based Approaches for Filtering Documents in a Parliamentary Setting

(12:15) Víctor Martínez, Fernando Berzal and Juan-Carlos Cubero

Probabilistic local link prediction in complex networks

(12:35) Fernando Bobillo, Lacramioara Dranca, and Jorge Bernad

A Fuzzy Ontology-based System for Gait Recognition using Kinect Sensor

(12:55) Sébastien Harispe, Massissilia Medjkoune and Jacky Montmain

Eliciting Implicit Evocations using Word Embeddings and Knowledge Representation


Tutorials 1 (October, 4 - 14:30-16:00)

Chair: Serafín Moral

(14:30) Nic Wilson (University College Cork, Ireland)

Preference Inference: Reasoning About User Preferences in a Combinatorial Setting

(15:15) Antonio Salmerón (University of Almería, Spain)

Analysis of massive data streams using probabilistic graphical models


Session 2 (October, 4 - 16:30-18:30)

Chair: Manuel Gómez-Olmedo

(16:30) Vu-Linh Nguyen, Sébastien Destercke and Marie-Hélène Masson

K-nearest neighbour classification for interval-valued data

(16:50) Pauline Minary, Frédéric Pichon, David Mercier, Eric Lefevre and Benjamin Droit

Evidential joint calibration of binary SVM classifiers using logistic regression

(17:10) Andrés Cano, Manuel Gómez-Olmedo, and Serafín Moral

Estimating Conditional Probabilities by Mixtures of Low Order Conditional Distributions

(17:30) Francisco Henrique Otte Vieira de Faria, Fabio Gagliardi Cozman, and Denis Deratani Mauá

Closed-Form Solutions in Learning Probabilistic Logic Programs by Exact Score Maximization

(17:50) Yang Xiang

Fault Tolerant Direct NAT Structure Extraction from Pairwise Causal Interaction Patterns

(18:10) Richard Billingsley, John Billingsley, Peter Gärdenfors, Pavlos Peppas, Henri Prade, David Skillicorn, and Mary-Anne Williams

The Altruistic Robot: do what I want, not just what I say


Welcome cocktail (October, 4 - 20:00-22:00)


Chair: Daniel Sánchez

Keynote 2 (October, 5 - 09:00-10:00)

Alberto Bugarín, University of Santiago de Compostela, Spain

What uncertainty models do we need for Natural Language in Data-To-Text Systems?

Automatic data-to-text (D2T) systems are increasingly being used to remove the barrier between data stored in information systems and the people that demand information from them. D2T are a part of the wide .eld of Natural Language Generation, which automatically generate high-quality narratives which summarize in natural language the most relevant information hidden in the data, which are typically numerical (as in time series), symbolic or both. The produced narratives are provided in a form that can be directly consumed and understood by human users. D2T systems are usually built as a way to directly information to users or as a complement of other means (graphical representations, numeric tables, . . . ). In this talk we will present and describe the general architecture of D2T systems (the so called D2T pipeline). We will also carry out a review of real D2T systems of high impact, in areas such as robot-journalism, meteorological forecasting, and clinical and industrial supervision systems, among others. We will present the models used in D2T for managing uncertainty in the language realizations. Special attention will be devoted to discuss the role that fuzzy logic and Computing with words play in the D2T .eld as well as to present potential areas of convergence between these two paradigms. Finally, we will present some open issues in D2T related to uncertainty management in natural language and to scalability of this approach within the contex provided by large volumes of data.


Tutorials 2 (October, 5 - 10:00-10:45)

Chair: Daniel Sánchez

Laurent Vercouter (LITIS/ INSA Rouen, France)

Community Detection and Clustering in Social Networks


Session 3 (October, 5 - 11:15-13:15)

Chair: Leila Amgoud

(11:15) Nahla Ben Amor, Didier Dubois, Héla Gouider, and Henri Prade

Expressivity of possibilistic preference networks with constraints

(11:35) Toshiko Wakaki

Assumption-Based Argumentation Equipped with Preferences and Constraints

(11:55) Sylvie Doutre and Jean-Guy Mailly

Semantic Change and Extension Enforcement in Abstract Argumentation

(12:15) Leila Amgoud and Jonathan Ben-Naim

Measuring Disagreement in Argumentation Graphs

(12:35) Sylwia Polberg, Anthony Hunter and Matthias Thimm

Belief in Attacks in Epistemic Probabilistic Argumentation

(12:55) Elise Bonzon, Jérôme Delobelle, Sébastien Konieczny and Nicolas Maudet

A Parametrized Ranking-based Semantics for Persuasion


Tutorials 3 (October, 5 - 14:30-16:00)

Chair: Fernando Bobillo

(14:30) Rafael Peñaloza-Nyssen (Free University of Bozen-Bolzano, Italy)

Fuzzy Description Logics

(15:15) Régis Sabbadin (INRA,France)

Exact and Approximate Inference in Graphical Models


Cultural Time (October, 5 - 16:00-20:30)

Visit to one of the most beautiful monuments in the world: The Alhambra.


Gala Dinner (October, 5 - 20:30-23:30)


Keynote 3 (October, 6 - 09:00-10:00)

Chair: Henri Prade

Martin Theobald, University of Luxembourg, Luxembourg

Scalable RDF Data Management with a Touch of Uncertainty

The invited talk provides an overview of our recent research activities and also highlights a number of research challenges in the context of extracting, indexing, and querying large collections of RDF data. A core part of our work focuses on handling uncertain facts obtained from various information-extraction techniques, where we aim to develop efficient algorithms for querying the resulting uncertain RDF knowledge base with the help of a probabilistic database. A further, very recent research focus lies on scaling out these approaches to a distributed setting. Here, we aim to process declarative queries, posed in either SQL or logical query languages such as Datalog, via a proprietary, asynchronous communication protocol based on the Message Passing Interface. Our current RDF engine, coined “TriAD”, has proven to be one of the fasted such engines over a number of RDF benchmarks with up to 1.8 billion triples.


Session 4 (October, 6 - 10:30-11:50)

Chair: Xiang Yang

(10:30) Steven Prestwich, Federico Toffano and Nic Wilson

A Probabilistic Programming Language for Influence Diagrams

(10:50) Marco Wilhelm, Marc Finthammer, Gabriele Kern-Isberner and Christoph Beierle

First-Order Typed Model Counting for Probabilistic Conditional Reasoning at Maximum Entropy

(11:10) Rafael Peñaloza and Nico Potyka

Towards Statistical Reasoning in Description Logics over Finite Domains

(11:30) Carlos Molina, Belén Prados-Suárez, Antonio Cortes-Romero

Bankruptcy Scenario Query: B-SQ


Tutorials 4 (October, 6 - 11:50-13:20)

Chair: M. Dolores Ruiz

(11:50) Inés Couso (University of Oviedo, Spain)

Maximum likelihood estimation and coarse data

(12:35) Olivier Cailloux (Laboratoire Génie Industriel, École Centrale Paris, France)

Reasons and Means to Model Preferences as Incomplete


Session 5 (October, 6 - 14:30-15:50)

Chair: Andrea G.B. Tettamanzi

(14:30) Nicolás Madrid and Manuel Ojeda-Aciego

A view of f-indexes of inclusion under different axiomatic definitions of fuzzy inclusion

(14:50) Carlos Molina, María D. Ruiz, Daniel Sánchez, and José M. Serrano

Ad hoc metric for correspondences analysis between fuzzy partitions

(15:10) Salem Benferhat, Khaoula Boutouhami, Faiza Khellaf and Farid Nouioua

An integer 0-1 linear programming approach for computing inconsistency degree in product-based possibilistic DL-Lite

(15:30) Ginés Moreno, Jaime Penabad and José Antonio Riaza Valverde

On Similarity-based Unfolding


Tutorials 5 (October, 6 - 15:50-16:35)

Chair: Daniel Sánchez

Francisco Herrera

University of Granada, Spain


Session 6 (October, 6 - 17:00-18:20)

Chair: Nic Wilson

(17:00) Laurent Garcia, Claire Lefèvre, Odile Papini, Igor Stéphan and Würbel Eric

A Semantic characterization for ASP base revision

(17:20) Sébastien Konieczny and Ramón Pino Pérez

On Iterated Contraction: syntactic characterization, representation theorem and limitations of the Levi identity

(17:40) Célia da Costa Pereira, Didier Dubois, Henri Prade, and Andrea G. B.Tettamanzi

Handling Topical Metadata Regarding the Validity and Completenes of Multiple-Source Information: A possibilistic approach

(18:00) Christophe Labreuche

Aggregation of preferences on criteria importance expressed on various subsets by several decision makers