Elena Bellodi and Riccardo Zese
Probabilistic Knowledge Representation in Machine Learning
Representing uncertain information is crucial for modeling current real-world domains. This has been fully recognized both in the field of Logic Programming and of Description Logics (DLs), with the recent introduction of Probabilistic Logic Languages in logic and with various probabilistic extensions of DLs respectively.
Machine learning approaches based on the combination of logic and probability have originated the field of Statistical Relational Artificial Intelligence (StarAI), which is getting an increasing attention.
In fact, probabilistic languages based on Logic Programming (LP) are particularly promising because of the large body of techniques for inference and learning developed in LP. On the other hand, Probabilistic Description Logics (PDL) can meet the Semantic Web need of representing and reasoning on structured but often incomplete data by exploiting knowledge representation formalisms that possess nice computational properties such as decidability and/or low complexity.
Battista Biggio and Fabio Roli
Adversarial Machine Learning
This tutorial aims to introduce the fundamentals of adversarial machine learning to the computer vision community, presenting a well-structured review of recently-proposed techniques to assess the vulnerability of machine-learning algorithms to adversarial attacks (both at training and test time), and some of the most effective countermeasures proposed to date. We discuss these issues also in the context of clear application examples including object recognition in images, biometric identity recognition, spam and malware detection.
Daniele Porello, Giancarlo Guizzardi and Nicola Guarino
Formal Ontological Analysis and Knowledge Representation
Formal ontologies are increasingly used in a variety of domains in crucial applications of AI, Knowledge Representation, Multiagent Systems, Conceptual Modelling, and Software Engineering. Ontologies are a way to express the information about a certain domain in a peculiar way: they intend to make the modelling choices and the assumptions of the modeler clear, justified, and sharable among the community of users.
This tutorial provides an introduction to the main foundational ontologies focusing on the motivations of the foundational choices and on the formal language used for developing the foundational ontologies. This tutorial is addressed to all the practitioners of AI whose work requires a grounded description of a particular domain and is intended to provide a number of tools to assess the quality of the modellisation.
Marta Cialdea Mayer, Nicola Gigante, Angelo Montanari, Andrea Orlandini and Alessandro Umbrico
Timeline-based Planning and Execution: Theory and Practice
Among the various approaches to planning and scheduling proposed in the literature, the timeline-based one proved itself quite successful in its deployment in a number of concrete applications, such as, for instance, autonomous space systems.
Despite its practical relevance, for a long time the theoretical properties of timeline-based planning have not been systematically investigated. In particular, a general picture of its computational complexity and expressiveness was missing until very recently.
The tutorial aims at providing a general overview of timeline-based planning and at giving an account of its formal properties. More precisely, it provides a complete formalization of the timeline-based planning, including flexible timelines and controllability issues in in the presence of uncertainty, and it analyzes its complexity and expressiveness (in comparison to action-based planning). Moreover, it gives a short introduction to the usage of a concrete timeline-based planning and execution system, called PLATINUm, that implements such a formal framework.
Diego Calvanese, Benjamin Cogrel and Guohui Xiao
Novel Developments in Ontology-Based Data Access and Integration (NOBDI)
Ontology-Based Data Access and Integration (OBDA/I) is a popular paradigm for overcoming the typical difficulties in accessing and integrating data stored in different kinds of legacy sources, by leveraging a conceptual representation of such data provided in terms of an ontology. To do so, it combines technologies developed in the areas of Knowledge Representation and in the Semantic Web with traditional database systems. The main mechanism for establishing such a combination are RDF-to-database mappings, which allow one to create virtual views characterizing the content of the data sources and to directly relate them to the terms in the ontology. In this tutorial: (i) We provide a general introduction to the principles and basic technologies for OBDA/I, relying on standard languages used in the Semantic Web, such as RDF, the OWL~2~QL ontology language, and the SPARQL query language. (ii) We provide more insights into the theoretical foundations of OBDA/I, by analyzing the approach based on query rewriting and the computational impact of ontologies and mappings on query answering. (iii) We provide an overview on the some recent advancements in OBDA/I concerning non-relational (NoSQL) data sources and the access to cross-linked data sources.