Research projects 2018-19

In the following the list of project offered for the next academic here is shown. Foreign candites applying to the second call can propose a project using a title either from the first or second call list. Candidates that are applying but don't need scholarship must refer to the list of projects of the first call.

Research projects for academic year 2018-19 (cycle 34 - second call)

 

Deep Learning end-to-end systems applied to Natural Language Processing	Prof. Rossella Cancelliere, apprenticeship contract by Loquendo S.p.A	Deep learning end-to-end systems are the cutting edge of research and technology challenge for machine learning. For example, they are applied with success to improve state of the art performance in Image and Speech processing. The focus of this Research Project is to investigate the different approaches to deep learning end-to-end modeling for the speech processing topic, and identify the most promising, possibly identifying solutions for improving the performance given specific domains (like for example language, etc.). The investigation is not limited to basic speech recognition step, but can be extended also to other fundamental stages like natural language processing etc. Considering the high availability of Neural building packages, like for example PyTorch, Tensor Flow, etc. the implementation work of the research activity could be based on such tools.
Progettazione e sviluppo di chatbot	Prof. Luigi Di Caro, apprenticeship contract by CELI s.r.l.	L'attivita' di ricerca si inserisce nel progetto di ricerca "CANP - la CAsa Nel Parco" finanziato attraverso il bando Piattaforma Tecnologica Salute e Benessere della Regione Piemonte , di cui CELl e UNITO sono partner. Il dottorato e' dedicato alla progettazione e sviluppo di chatbot (sistemi di dialogo) per la progettazione partiecipata e l'assistenza domiciliare ai pazienti.
Video Ispezione di Infrastruttura Ferroviaria con l’ausilio di Reti Neurali / Deep Learning for Visual Inspection of Railways Infrastructure	Prof. Marco Grangetto, scholarship funded by DMA	Il progetto di ricerca prevede lo studio e lo sviluppo di un sistema di visione automatica per riconoscimento della segnaletica in ambiente ferroviario e per la localizzazione e/o classificazione delle varie componenti del binario basato su progettazione e addestramento di una o più reti neurali; il progetto dovrà inoltre tenere conto dei requisiti di analisi in tempo reale richiesti dall’applicazione valutando la compatibilità dell’esecuzione della stessa su piattaforme di calcolo con capacità diverse, ad esempio PC/workstation e sistemi di tipo embedded.

 

Research projects for academic year 2018-19 (cycle 34 - first call)

 

Processi dinamici e analisi strutturali in reti complesse/ dynamical processes and structural analyses in complex networks	Prof. Giancarlo Ruffo	Nowdays data scientists deals with the understanding and forecasting of real world phenonema adopting a data-driven approach to the analysis of complex systems. One of the most successful representations that allowed researchers to uncover non trivial properties from large data sets is the network (or graph) that have led to the breakdown of standard theoretical frameworks and models. In fact, a vast number of real world systems, from the socio-economic domain to power grids and the Internet, can be represented as large scale complex networks. Measuring structural characteristics and the dynamics of a network may lead to new level of undersanding of complex processes such as virality of information, diffusion, synchronization, and resilience. New computational and algorithmic challanges arise to deal with large datasets, and novel applications must be designed according to natural self organization of such complex systems.
Crowdmapping/Crowdmapping	prof. Guido Boella	Examining the impact of IT on the evolution of participatory cartography into crowdmapping. Possible focuses are: georeference by means of GPS (outdoor maps) or RFID (indoor mapping) in order to collect environment and user data, integration of map semiotics with user interaction principles, analysis of the social aspect by means of social networks and incentive mechanisms, cognitive (and possibly ontology-based) models for map-based knowledge representation, analysis of the temporal dimension for the map-based representation of fluents. Projects: FirstLife
Applicazioni di blockchain e smart contract/Blockchain applications and smart contracts	prof. Claudio Schifanella	The success of Bitcoin chryptocurrency has shown the potentialities of the distributed ledger technologies such as the blockchain. This project aims at studying these technologies and in particular of smart contracts. This project will benefit from the connection with the Co-city European project, where a blockchain for social purposes is developed
Modelli di computazione in memoria e loro applicazioni al Deep Learning e al Big Data Analytics/Near Data Processing and its applications to Deep Learning and Big Data Analytics	prof. Marco Aldinucci	The research aim at defining a novel distributed memory abstraction to support Near Data Processing (NDP). The NDP-enabled memory will be envisioned to be programmable by way of a novel DSL and to exhibit a well defined set of invariants among computation epochs. Also, it should be able to support distributed scalable pipelines for BigData and Machine Learning pipelines.
Supporti a tempo di esecuzione per applicazioni su stream nel paradigma di computazione edge/Run-time supports for stream processing on edge computing.	prof. Marco Aldinucci	Recently, Internet-of-Things applications started to address scenarios where data inputs consist of highly dimensional data coming from multiple sources and/or characterized by different feature subsets. For example, a person can be identified by face, finger-print, EEG brainwaves and irises, each coming from a different sensor, while the surface of a physical object can be represented by its color and texture attributes. These data streams should be computed in-place and on-line (in the edge); they cannot simply be moved to a data lake because of both functional (e.g. privacy) and extra-functional (e.g. performance) requirements. The reserach will focus on methods and tools to orchestrate and program large ecosystems of complex streamign devices in the edge.
Calcolo ad altre prestazioni per la fisica delle alte energie/High-Performance Computing for High Energy Physics	prof. Marco Aldinucci, prof. Massimo Masera	The ALICE@CERN experiment, after the first 9 years of operations, will be equipped with a new Inner Tracking System (ITS) detector during the 2-years long LHC shutdown scheduled for 2019-2020. This detector will feature seven layers of silicon pixel sensors for a total of 12 Giga-pixels and it will be able to successfully detect up to 50k lead ions interactions per second that leave a signal in the seven layers constituting the ITS. These signals collected every 1-5 ms produce a data throughput of approximately 40 GB/s that should be processed online at the experiment site and reconstructed as thousands of helicoidal trajectories of charged particles moving in a magnetic field (a.k.a. tracking). In the ALICE@CERN experiment the tracking problem has been traditionally faced by way of shared-memory multithreading, which hardly scale to to the throughput required by the new ITS. The project will face the problem of designing a novel class of data parallel tracking algorithm based on the Cellular Automata model and their engineering on the SIMT/GPU computing paradigm.
HRI - Human Robot Interaction for social, assitive, and educational purposes	prof. Cristina Gena	Socially interactive robots are becoming now a reality, since commercial social robots are now already available in real contexts not just for experimental purposes, and they will always be more present in our daily life. However, social robots are a new species that must be studied, tested and redesigned according to the HCI methodologies. Social robots are complex cognitive and physical artefacts that requite interdisciplinary Human Robot Interaction team having competence that range from AI (machine learning, natural language processing, knowledge representation, planning, etc.) to HCI (interaction design, affective computing, emotion recognition, naturual language interaction, multimodal interaction, etc). We are now starting to work in this complex area, both experimenting commercial social robots and designing new ones. We are particularly interested, but not only, in experimenting social robot as assistants in educational context (educational robotic), as therapeutic and social companion in assistive scenarios (autism, elderly people, etc. ), etc.
BCI - Brain Computer Interaction in the context of smart and IoT environements	prof. Cristina Gena	Brain Computer Interfaces (BCI), are interfaces that put the user in communication with an electronic device using electrical signals biologically produced by the human body, especially the brain waves. These technologies are now having an enormous growth and, in the future, will be able to restore autonomy to those who unfortunately lost the ability to move independently due to accidents or disabling illnesses. However BCI are also investigated in other contexts, such as games, military services, aeronautics, etc. In computer science, BCI benefits from contributions from HCI and AI We are particularly interested, but not only, in studying and experimenting BCI technologies for helping people with disabilities in a IoT context, for in instance in a smart home and/or smart environment, and also in conjunction with social and interactive robots that could help users in preforming tasks for them
Logics and Models of Innovative Computing Models	prof. Luca Paolini	Reversible computing is, on its own, an unconventional form of computing. It is a crucial aspect of many innovative emerging computational models as bio and quantum computing. It imposes of a backward determinism that allows us to reverse the computation so that, for example, we can undo a reversible program step by step re-establishing former situations. Fortunately nothing is loss, in the sense that all classical computations can be encoded into reversible ones. The quantum computing is one of the most promising generalization of the reversible model, but its relevance is expected to be disruptive also in classical computing with impacts on hardware and software. We aim to deepen our logical and formal knowledge about them, maybe looking to further unconentional computing models, as the probabilistic and the analof ones.
Designing intelligent behavior and interaction for smart objects	prof.Luca Console	Smart objects are very popular nowadays. However not all dimensions of intelligence have been explored by recent reserch on smart objects and many methodologies and solutions from teh last decades of research in Artificial Intelligence can be exploited to improve intelligent behavior and intercation in everyday objects. we aim at studying these aspects ypwards intelligent spaces and ecosystems
Mining, retrieval e analisi di processi di business /Mining, retrieval and analysis of business process models	prof. Stefania Montani	Process mining techniques to learn process models from business process traces; definition of proper similarity metrics for business process models; exploitation of these metrics within proper retrieval and ordering algorithms to support process analysis; testing in real world domains (e.g. stroke management; optimization of process task scheduling in a cloud computing environment).
Tecniche di case-based retrieval flessibile/Flexible case-based retrieval techniques	Prof. Stefania Montani	Representation of cases with time series features; dimensionality reduction techniques; efficient retrieval techniques also with partial match between the query and the retrieved cases; indexing and retrieval optimization techniques, also in sinergy with deep learning approaches. Application to medical and bioinformatics problems
Tecniche di intelligenza artificiale per l'informatica medica/Intelligent techniques in medical informatics	Paolo Terenziani and Stefania Montani	The project aims at developing intelligent techniques to support physicians in the treatment of patients through clinical guidelines. The work is part of the GLARE project, a project for the development of a software prototype for the acquisition, representation and execution of clinical guidelines.
Computer vision and deep learning for multi-dimensional imaging and modeling/ Visione artificiale e deep learning per immagini multi-dimensionali	prof. Marco Grangetto, Maurizio Lucenteforte	The way we capture visual information from reality is changing significantly as witnessed by technological development of imaging sensors, ranging from stereoscopic/spherical to depth sensors, and light-field cameras. On the one hand, the imaging sensors can support novel display technologies, e.g. HDR, auto-stereoscopic, holographic and immersive head-mounted displays. On the other hand, new multidimensional images (including RGB, infra-red, depth, 360-degrees), coupled with the advance of computer vision based on deep learning models pave the way to unparalleled opportunities for understanding and modeling of visual information. This project will be focused on advancing the state of the art in the area of computer vision with potential impact in many application areas ranging from image processing to 3D rendering and augmented reality. http://di.unito.it/eidoslab
Apprendimento nelle metaeuristiche di ottimizzazione / Learning in Metaheuristic Optimization	Prof. Andrea Grosso, Roberto Aringhieri	The power of modern metaheuristic algorithms to generate better solutions to complex optimization problems is based on the link between algorithmic effectiveness and the combined effects of intensification, diversification and learning. This proposal aims at exploring the introduction of deep learning tools in the development of metaheuristic algorithms for hard combinatorial problems arising in the scheduling, rostering and timetabling fields.
Analisi e sviluppo di politiche per la gestione di una rete di servizi sanitari basate su Big Data /Big Data supporting health care network policies	Roberto Aringhieri, Andrea Grosso	Health Care Big Data (HCBD) enables a detailed health system analysis: exploiting the HCBD, one can replicate the behaviour of the system modelling how each single patient flows within her/his care pathway. This proposal aims at providing an innovative and comprehensive methodology (mainly based on simulation and optimization) for the analysis of a health care system, which exploits the HCBD in such a way to improve both its efficiency and the equity in the access.
Apprendimento di reti neurali profonde per il trattamento di immagini e di sequenze spazio-temporali/Deep learning for image and spatio-temporal sequences processing	prof. Rossella Cancelliere	The proposal concerns the development and the experimental evaluation of deep neural architectures. Deep learning achieves great power and flexibility by learning to represent the world as a nested hierarchy of concepts, with each concept defined in relation to simpler concepts, and more abstract representations computed in terms of less abstract ones. The project aims at investigating these features, in order to further improve performance in classical tasks like classification and diagnosis, and also in image and speech processing and data to text generation, where the spatio-temporal continuity plays a fundamental role in defining the final categories.
Sistemi avanzati di Open Information Extraction basati su tecniche di Natural Language Processing, Machine Learning ed integrazione di risorse semantiche / Advanced Open Information Extraction systems based on Natural Language Processing, Machine Learning and integration of semantic resources	prof. Luigi Di Caro	Open Information Extraction (OIE) was born to solve the problems of Information Extraction which does not scale well with large corpora and sets of possible semantic relations. The project involves tasks such as 1) the advancement of extracting methods, 2) the creation of multilingual datasets for applying Machine (and Deep) Learning algorithm and 3) the use of OIE for applications such as chatbots
Interfacce utente intelligenti per disabilità cognitive/Intelligent user interfaces for cognitive disabilities	Federica Cena	The aim of the project is to study, design and develop intelligent solutions for people with cognitive disabilities (Autistic Spectrum Disorder, Dementia, etc.) A special focus will be devoted to interactive spatial services of urban environments for social inclusion.
Sensors and web minining for Personalized persuasive technologies	Federica Cena	The aim of the project is to exploit heterogeneous personal data coming from different sources (werables, ubiquitous sensors as well as web) in order to design novel personalised services to change opinions and behaviour
Reti neurali psicologicamente plausibili per l'apprendimento linguistico/Psychologically plausible neural networks for early word learning	Valentina Gliozzi
Mixing Deep Learning and symbolic reasoning for Natural language generation (NLG)	Alessandro Mazzei	Deep neural networks (DNN) have only recently gained attention from Natural Language Generation (NLG) research community. Indeed, traditional symbolic architecture based on a strict decomposition of the generation processes in distinct phased have the big advantage to be self-explaining for human beings. In this research project we want to investigate on the possibility to unify the predictive power of DNN with the explainable capabilities of symbolic NLG systems by using the former as an "oracle" for the latter.
Modeling and analysis of fake/polluted information generation and diffusion over social media	Rossano Gaeta - Michele Garetto	The research project aims at the definition of simple mathematical models to describe and analyze the generation and diffusion of fake/polluted information over social media (e.g., Facebook, Twitter, etc). The activity will start from the analysis of available real data on fake/polluted information diffusion on online social networks whose results will be exploited for the development of analytical models. Simulation of detailed diffusion mechanisms will be used to validate model results.
Machine unlearning: protecting user privacy by making intelligent systems forget	Ruggero G. Pensa	Today's systems produce a rapidly exploding amount of data, and the data further derives more data, forming a complex data propagation network. There are many reasons that users want systems to forget certain data including its effects on intelligent systems and models. From a privacy perspective, users who become concerned with new privacy risks of a system often want the system to forget their data and the derived effects on machine learning models. From a security perspective, if an attacker pollutes an intelligent system by injecting manually crafted data into the training data set, the system must forget the injected data to regain security. The project concerns the study of forgetting systems, capable of forgetting certain data and their effects on trained models, completely and quickly. It focuses on making learning systems forget, a process of which called machine unlearning, or simply unlearning. Existing approaches transform learning algorithms used by a system into a summation form. To forget a training data sample, they simply update a small number of summations. However, not all models can be reduced to summation forms or their reduction is not straightforward. This project focus on the theoretical and experimental extension of machine unlearning to more realistic and complex learning scenarios.
Differentially private mechanisms for co-clustering algorithms	Ruggero G. Pensa	Privacy-preserving machine learning is the ability of learning from data without disclosing any private information of the users. Among all proposed approaches, differential privacy has gained popularity due to its effectiveness and robustness in many application domain, including in clustering algorithms. Differentially private co-clustering, however, has never been defined. Thanks to its ability in partitioning data objects (e.g., customers) and properties (e.g., products) simultaneously, co-clustering enables the discovery of self-described clusters. Cluster descriptions, however, introduce additional and potentially harmful information that may support the disclosure of users' preferences. This project concerns the study, design, implementation and experimental validation of differentially private co-clustering algorithms through the adoption of Laplacian and/or exponential mechanisms.
Model checking quantitativo/Quantitative Model Checking	Jeremy Sproston	Model checking is a system verification technique for establishing whether a model of a system satisfies a number of formally-specified correctness properties. Model checking has been extended so that quantitative information, for example regarding continuous, timed or probabilistic behavior, can be represented both in system models and in the correctness properties. The project concerns the development of model-checking techniques for quantitative aspects of embedded software and cyber-physical systems in general, with particular focus on timed and probabilistic issues
Ingegnerizzazione Rigorosa del Software per l'Internet degli Oggetti /Rigorous Software Engineering for the Internet of Things	prof. Ferruccio Damiani	The Internet of Things is ushering a dramatic increase in number and variety of interconnected smart devices. Inherent distribution, mobility, situatedness, and heterogeneity of such devices calls for proper scientific understanding of the foundations of such systems as well as for novel software methods. The project investigates formal aspects of software engineering for Aggregate Computing, a promising approach for constructing a convenient abstraction layer on top of large scale ubiquitous systems. It is based on the idea of shifting from the standard single-device focus to an aggregate viewpoint in which one sees the overall set of computational devices spread in the pervasive computing environment as a single "machine", a sort of diffused computational fabric. [di.unito.it/movere/]
Metodi Formali per le Linee di Prodotti Software/Formal Metods for Sofeware Product Lines	prof. Ferruccio Damiani	Modern software systems are often built from customizable and inter-dependent components. Such customizations usually define which features are offered by the components, and may depend on backend components being configured in a specific way. As such system become very large, with a huge number of possible configurations and complex dependencies between components, maintenance and ensuring the consistency of such systems is a challenge. Software Product Line Engineering (SPLE) is an established approach to address this challenge. A Software Product Line (SPL) is a family of similar programs generated from a common artifact base. A Multi SPL (MPL) is a set of interdependent SPLs that are typically managed and developed in a decentralized fashion. Delta-Oriented Programming (DOP) is a flexible and modular approach to implement SPLs. This project investigates new concepts that extend DOP to support the implementation of MPLs. [di.unito.it/movere]
Recupero di informazioni geografiche personalizzato / Personalized Geographic Information Retrieval	Prof.ssa Liliana Ardissono	Collaborative Geographic Information Systems support information sharing in a one-size-fits-all way, i.e., the same search query receives the same information, regardless of the user submitting it. However, information needs may differ as a consequence of different search contexts (e.g., for work, for the family), and different user preferences/interests. Starting from the research work carried out in project OnToMap (URL: https://ontomap.ontomap.eu/), models and techniques for personalizing geographical information filtering and presentation, on the basis of the search context and of individual user profiles, will be defined and tested by exploiting large-scale search logs. The developed models will be applied to the OnToMap Participatory GIS, which supports information retrieval with map visualization.
Explainable dynamic constraint reasoning	Luca Anselma - Alessandro Mazzei	Constraint based reasoning can be used to formalize and resolve different kinds of problems. For instance, Simple Temporal Problems (STPs) have been applied for dynamic problems related to the quantified-self, e.g. diet management. However, these systems are often not self-explainable, i.e. they require external resources to explicate the results of the numeric reasoning. In this project we want to investigate the possibility to use symbolic external knowledge (e.g. ontologies) to explain, with Natural Language Generation, the causes and the consequences of the STP reasoning.
Theory and practice of concurrent programming languages	Luca Padovani	Modern software applications running on distributed and multi-core computing architectures heavily rely on concurrency, inter-process synchronization and communication. These constructs are challenging to reason about and lead to subtle programming mistakes that are difficult to track and to repair. As a result, the overall development process of these applications is more expensive and less productive. This PhD project aims at developing type-based theories, techniques and tools for programming concurrent applications with certified correctness properties such as communication safety, protocol conformance, data-race freedom and deadlock freedom. The project will leverage on the extensive literature on behavioral types and logics for concurrency and will cover all aspects of concurrent program development, from static analysis to type-based runtime monitoring. Particular emphasis will be given to the application of the developed techniques and tools to real-world functional and object-oriented languages.
Advanced Methodologies for Temporal Relational Databases	Paolo Terenziani	Thirty years of research have widely demonstrated that the treatment of temporal information in the relational context requires specialized techniques. While the treatment of single facts that occurred at exactly known time is supported by many approaches, the treatment of repeated, periodic, and\or temporally indeterminate facts is still an open research problem. The PhD project aims at facing such a challenginf problem, with specific emphasis on the treatment of temporally indeterminate facts, i.e., of facts whose exact time of occurrence is not known (and can only be approximated). The goal of the PhD project is to analyse the semantics of such data, to  identify new data models to represent them, and to propose new relational algebrae to query them. Additionally, the properties of such new approaches (both theoretical and computational) will be analysed.
Ragionamento su azioni e ontologie/Reasoning about actions and ontologies	Laura Giordano, Daniele Theseider Dupré	Combining action languages with ontologies has been shown to be useful to model business processes as well as clinical guidelines while representing the ontological relations about concepts in the domain. The project will explore these relationship, considering in particular the combination of rule based languages and low complexity description logics of the EL and DL families, for which reasoning services, such as subsumption and instance checking, are polynomial.
Accountability computazionale/Computational accountability	Matteo Baldoni, Cristina Baroglio, Roberto Micalizio	Business processes represent an important tool for rationalizing business resources and cross-business relationships. Business processes are mainly based on a procedural and prescriptive representation of activities. Their main disadvantages are: the concerned parties cannot take advantage from occasions nor adapt to adverse situations; there is no representation of accountability; they are not suitable to realize socio-technical systems. The proposed research aims at studying social relationships as an alternative for realizing business processes. Social relationships capture dependencies among the actors, capture in a natural way a notion of accountability, and they evolve exclusively after the observation of social behaviour. Due to their declarative nature they are suitable to describe business processes in a minimally descriptive way.
Interazione e coordinazione di sistemi multiagente basata su relazioni sociali /Interaction and coordination based on social relationships for Multiagent Systems	Matteo Baldoni, Prof. Cristina Baroglio	Commitments and, more generally, social relationships, are the emerging paradigm for modeling interactions and coordination in multiagent systems. This research will investigate the use of social relationships both form a theoretical and practical point of view. A special attention will be posed on data-awareness and norm-awareness.
Tecnologie semantiche e sentiment analysis per la valorizzazione dei beni culturali / Semantic Technologies and sentiment analysis to enhance the value of cultural heritage	Rossana Damiano, Viviana Patti, Anna Goy	The project can be seen as part of the inter-disciplinary research area of Digital Humanities, and aims at studying, designing, and developing ICT-based solutions for cultural resources management, with the objective of making them accessible and usable in innovative ways by a large audience with different skills and interests. In order to achieve this goal, the project will investigate the integration of two perspectives: (a) Using semantic technologies (ontologies, Semantic Web resources, Linked Open Data) to represent the content of heterogeneous cultural resources (textual documents, pictures, videos, paints, sculptures, monuments, etc.). The resulting conceptual semantic model should enable the identification of people, places, events and relationships, thus allowing users to "discover" alternative and original access paths to cultural heritage. In particular, narrative paths plays a major role to support a direct and friendly communication with users. Special attention will be payed to the possibilities offered by crowdsourcing models, in order to collaboratively build semantic representations of the domain and of itsinterpretation. (b) Using sentiment analysis tools. Such tools have proven useful to extract opinions and mood from a range of textual data in different domains, such as media marketing, political and social analysis, etc. Aspects related to sentiment are also linked to aesthetic experience, as acknowledged by philosophical and psychological theories along centuries. Through social media, linguistic feedback about artworks or other types of cultural resources becomes available, thus the sentiment induced by such resources can be investigated. The goal of the PhD project is to investigate in a systematic way the relationship between sentiment, emotions, and content semantic description, in the field of cultural heritage, being artworks or archival resources. The research is expected to deliver a model of the relation between cultural resources and sentiment (or mood), by identifying domain-specific categories for describing sentiment in cultural heritage and its sub-domains. Coupling sentiment analysis and content semantic description in cultural heritage paves the way to innovative applications, that range from sentiment mood-based content recommendation, to personalized tools for the exploration of historical and art archives.
Smart engaging interactions with chat-bot	Rossana Damiano, Alessandro Mazzei	There are two main traditions in the architectures of dialog systems. The traditional one formalizes a structured notion of interaction, while recent approaches focus on the retrieval of the appropriate response based on large scale corpora. The idea of this project is to merge these two lines of research by using the paradigm of social and emotional management of the interaction. The goal is to design and implement multimodal interactive systems which engage the users in a natural interaction in heritage domains.
Tecnologie di Trattamento Automatico del Linguaggio Naturale per Sentiment Analysis ed Opinion Mining / Natural Natural Language Processing for Sentiment Analysis and Opinion Mining	Cristina Bosco	Provided the widespread interest for Sentiment Analysis and Opinion Mining, the project is oriented to the investigation of the contribution that computational linguistics can give to the development of systems for the tasks of this area, such the detection of stance, irony, polarity or hate speech. For instance, considering the recent advancement in morphological and syntactic representation of texts from social media, a novel direction can be followed which encompasses parsing technologies and sentiment analysis, also taking into account a multi-lingual perspective in order to improve the performance of systems and to shed some light on the deeper motivations of the success and failure of Sentimen Analysis techniques. As possibile scenarios where the research can be applied we propose politics (like the study of political debates) or social context (the study of phenomena related to hate speech against immigrants in European society).
Logiche descrittive preferenziali per la revisione di ontologie /Preferential Description Logics for Ontology Revision	Roberto Micalizio e Gian Luca Pozzato	The main goal of this project is to develop a methodology to revise a Description Logic knowledge base when detecting exceptions. The starting point of this approach relies on the methodology for debugging a Description Logic terminology, addressing the problem of diagnosing incoherent ontologies by identifying a minimal subset of axioms responsible for an inconsistency. Once the source of the inconsistency has been localized, the identified axioms are revised in order to obtain a consistent knowledge base including the detected exception. To this aim, we intend to adopt a nonmonotonic extension of the Description Logics based on the combination of a typicality operator and the well established nonmonotonic mechanism of rational closure, which allows to deal with prototypical properties and defeasible inheritance.
Logiche descrittive probabilistiche per la combinazione di concetti/Probabilistic Description Logics for concept combination	Antonio Lieto e Gian Luca Pozzato	Inventing novel concepts by combining the typical knowledge of pre-existing ones is one the most creative cognitive abilities exhibited by humans. Dealing with this problem requires, from an AI perspective, the harmonization of two conflicting requirements that are hardly accommodated in symbolic systems: the need of a syntactic compositionality (typical of logical systems) and that one concerning the exhibition of typicality effects. The main aim of this project is to provide a logical framework able to account for this type of human-like concept combination based on a nonmonotonic Description Logic with a probabilistic semantics. The applications of such a logical framework may impact on a wide variety of fields ranging from the creation of new characters in the movie industry (where, for example, the creation of a new protagonist for a cartoon can be obtained by combining some typical features from previous characters in a novel, surprising, way) to the automatic generation of novel and cognitively-plausible landscape scenarios or narrative story-lines in videogames. The proposed logical framework can be generally thought as a problem solving heuristic applied by artificial agents (including robots) in any situation where there is the need of inventing a novel solution by leveraging from a repertoire of pre-existing knowledge to be recombined.
Modelli Computazionali di Cognitive Decision Making e loro applicazioni per lo sviluppo di Tecnologie Persuasive sul Web / Computational Cognitive Models of Decision Making and applications to Persuasive Technologies in the Web	Antonio Lieto	An increasing percentage of the world's population is now connected to the Web as a global information environment and new forms of social interactions are are radically changing how the information spreads. One of the main problems arising from this massive availability of textual information (in web and social media) is represented by the spread of misinformation and fallacious arguments (somehow purposely manipulated by companies of by intelligent agencies). The current state of the art argumentative technologies are inappropriate to deal with this problem since the mere presentation of logical counter-examples is not sufficient to stop the spreading of fallacious arguments. The goal of this project is to develop innovative technologies based on well-founded computational models of cognition and cognitive architectures that should be able to persuade users about the unsoundness of fallacious argument thus preventing their diffusion. This doctoral project will be carried out within an already existing international scientific collaboration with researchers from the University of Haifa (Department of Information System, Israel) and from Carnegie Mellon University (USA).
Cognitive Knowledge Representation Systems and Formalism for Common-Sense Reasoning	Antonio Lieto e Daniele P. Radicioni	Representing and reasoning with common-sense knowledge is one of the main open challenges in the AI research. While classical approaches deal with these problems by proposing non-monotonic extensions of logical formalisms (e.g. Description Logics), cognitive approaches adopt well founded cognitive frameworks aiming at providing efficient non-monotonic/defeasible inferences by taking into account a plethora of reasoning heuristics coming from the field of Cognitive Science. The goal of this project is to develop well-founded computational cognitive models of human categorization in order to extend the current categorisation limitations of artificial systems. This doctoral project will rely and extend previous systems developed at the University of Turin (http://www.dualpeccs.di.unito.it/).
Lexical resources for semantic analysis	Daniele P. Radicioni	The tremendous growth of the available text documents due to the spread of the Web and of social networks demands for novel tools and approaches to automatically elaborate documents, for grasping their semantics. In the last few years novel resources have been proposed (ConceptNet Numberbatch, NASARI, GloVes, COVER, etc.) that allow to extract information useful in various applicative domains, such as documents categorization, keywords extraction, question answering, text summarization, etc.. This doctoral project is aimed at providing artificial systems with human-level competence in understanding text documents by leveraging the notion of common-sense, which is typically missing from existing resources. This project requires and promotes a multidisciplinary perspective, including Computer Science, Cognitive Science, Cognitive Psychology and Neuroscience. Doctoral candidates will collaborate in carrying on research on lexical semantics, as described at http://ls.di.unito.it. Along with lexical resources and theoretical models of language semantics, applications will be drawn according to the interest of candidates, to tackle, e.g., figurative uses of language (such as metaphor and metonymy recognition), fake detection in social media, open domain question answering, events detection and extraction.
The New Science of Cities	Rossano Schifanella	Urban informatics uses the vast amount of data generated from diverse data sources, e.g., social media feeds, transit cameras, mobile phones, and myriad of sensors, to better understand how cities work. This understanding can remedy a wide range of issues affecting the everyday lives of citizens and the long-term health, sustainability, and livability of urban centers - from morning commutes to emergency preparedness to air quality. This research project combines methods from a wide range of disciplines, e.g., machine learning, network science, social science, and spatial analysis, to model the dynamics of modern cities and to design and experiment new digital services with the ultimate goal to improve citizens quality-of-life.
Big Data applied to societal challenges	Rossano Schifanella	Scientists increasingly have access to data sets of unparalleled scope and complexity. New technologies have made the collection of that data - by governments, private companies, or innovative researchers - possible. Advances in computer science and statistics have allowed for inferential, simulated, and visual analyses that are now being incorporated into our daily life through online services. This project aims at contributing to a new area of research, called Computational Social Science (CSS), which aims to analyze socio-cultural phenomena through large-scale digital datasets and novel technologies. For those new to the social sciences, the project is an opportunity to see where your computer science and statistical skills can go, with innovative applications to problems of massive societal interest. For those new to computational methods, this is a chance to develop the tools necessary to design services that have an impact in the real world at scale. The project tackles these tasks with approaches from machine learning, text and data mining, network analysis, and behavioural science. Particular attention will be drawn on important issues like data ethics, privacy and algorithmic trasparency.
Combined effect of content quality and social ties on user engagement	Rossano Schifanella	The dynamics of attention in social media tend to obey power laws. Attention concentrates on a relatively small number of popular items neglecting the vast majority of content produced by the crowd. Although popularity can be an indication of the perceived value of an item within its community, previous research has highlighted the gap between success and intrinsic quality. As a result, high quality content that receives low attention remains invisible and relegated to the long tail of the popularity distribution. Moreover, the production and consumption of content is influenced by the underlying social network connecting users by means of friendship or follower-followee relations. This inderdisciplinary project aims at studying the complex intertwinement between quality, popularity and social ties in online platforms, designing a methodology to democratize exposure and foster long term users engagement.
Trattamento delle eccezioni nelle Logiche Descrittive: un approccio multi-preferenze/Multipreferences for dealing with exceptions in Description Logics	Laura Giordano, Valentina Gliozzi	The treatment of exceptions in ontologies has received a lot of attention in the last decade. The project will deal with the preferential approach to non monotonic reasoning in description logics and, specifically, with the use of multipreferences to enrich the preferential semantics underlying rational closure in order to separately deal with the inheritance of different properties.
Emotions and Moral Values in Expressions of Hatred in Social Media	Rossana Damiano, Viviana Patti	Manifestations of social, cultural and political opinions in social media are acknowledgedly characterised by a strong affective component. In particular, hate speech has been pointed out as an extreme, yet typical, manifestation of the expression of opinions, with peculiar features which can be detected and analysed with automatic tools, thus helping to detect and monitor the spread and distribution of this phenomenon. However, current models don't account for the complex interplay of emotions, intended as cognitive and social constructs, and of their manifestation in language. Emotions, and negative emotions in particular, are characterised by a social nature, partly due to their intrinsic relations with the enforcement of the values of some society. As such, they possess a potential for advancing the analysis of sentiment in social media language, especially when cultural and political opinions are expressed. In order to fill this gap, this project aims at investigating:- the role of emotional states in the expression of opinions on highly controversial topics, with the help of the established models of emotions delivered by the literature in psychology and linguistics; - the relations of emotional states with the implicit and explicit expression of values, and their emergence in language, with the use of linguistic analysis tools which also can deal with the presence of figurative and rhetoric devices; - the relation between the expressed emotions, and moral emotions in particular, and the linguistic notion of sentiment. The aim of the work is to accommodate sentiment, emotions and values in unified model, tailored to social media language, which can be employed to understand, monitor and prevent the extreme manifestations of social, cultural and political debate.
Cooperazione nel fog computing: progettazione, modellazione ed analis/Cooperation in fog computing: design, modeling, and analysis	prof. Rossano Gaeta, Marco Grangetto	Recently, fog computing emerged as a possible architectural solution to drive a smooth convergence of clouds and mobile applications. The target of the research activity is the study, design, modeling, evaluation, and analysis of techniques to obtain effective and efficient logical organization of fog servers and to achieve high level of joint service delivery availability, performance, robustness, and privacy.
Propagazione dell'informazione su reti complesse/Information propagation on complex networks	Prof. Giancarlo Ruffo, Francesco Bonchi	The big data challenge, including the availability of huge volumes of information streaming from popular social media and microblogging platforms, allows the unprecedented opportunity of posing new scientific questions on a different scale. Sophisticated empirical analyses of a variety of social phenomena have now become possible thanks unprecedented availability of network data. Among these phenomena, one that has attracted a lion's share of interest is the study of social influence, i.e., the causal processes motivating the actions of a user to induce similar actions from her peers, creating a selective sweep of behavioral memes. Mastering the dynamics of social influence (i.e., observing, understanding, and measuring it) could pave the way for many important applications, among which the most prominent is viral marketing, i.e., the exploitation of social influence by letting adoption of new products to hitch-hike on a viral meme sweeping through a social influence network. The goal of this project is to develop new methods and applications for the analysis of information propagation over social networks and the phenomenon of social influence, using a multidisciplinary approach, conjugating machine learning, probabilistic causal theory, game theory and digital epidemiology. This project could be co-advised with Dr. Francesco Bonchi (ISI Foundation).
Agente intelligente per la comprensione automatica di domini scientifici ed il dialogo personalizzato via chatbot / User-Adapted Domain Knowledge Processing for supporting scientific text comprehension via chatbot	Luigi Di Caro	L'obiettivo del progetto è il design e la creazione di un agente intelligente in grado di costruirsi una base di conoscenza da corpora testuali su un dominio scientifico (ad es. vaccini) e sull'utente che vuole interagire (interfacciandosi anche a profili social quali Facebook e Twitter, ed effettuando stance/opinion mining). L'agente intelligente interagisce attraverso interfacce testuali a dialogo personalizzate (chatbot), cercando di favorire il confronto su aspetti contrastanti, con un linguaggio adattivo.
Mining for the social good: analysing data for the benefit of people	Rosa Meo	In the era of big data, open data that regards people and the communities are collected and made public. We are interested in analysing these data for the benefit of people. For instance, we are interested in analyzing the relationships between air pollution and breathing problems, analysing the data of traffic and the use of eco-friendly transportation, or possible cause-effect relationships between students life styles and their outcomes in studies or their professional choices. We will do our data mining without causing any harm to the people privacy, by employing privacy preserving techniques.
Industry 4.0: analysing data in the companies sensors	Rosa Meo	Companies have many sensors in their productive systems, measuring the conditions at which they operate. We want to analyse these data, how they correlate and evolve in time in order to improve production (in costs and quality), reduce the energy consumption and the overall emissions.
Data-driven decision making in complex systems	Maria Luisa Sapino	Data-driven decision making is increasingly critical in many application domains, including IoT and Smart Cities.Yet, several critical data challenges remain. Decision making in complex systems may require tracking of 10s or 100s of data sources, spanning multiple layers and spatial-temporal frames, affected by complex inter-dependent dynamic processes. Moreover, due to large number of unknowns, decision makers usually need to generate an ensemble of models, requiring 10s-1000s of alternative outcomes. Situation on the ground evolves unpredictably, requiring continuous adaptation of models and decisions. As data (encoded as multi-variate time series) flows through the system, it needs to be processed and indexed and the provenance of such data needs to be automatically recorded by the system, for later use and analysis. This research aims to create a data-and decision flow system, DataStorm, through efficient and scalable integration of big data and machine learning techniques to support real-time and accurate decision making.
Elaborazione di immagini biomediche per il supporto alla diagnosi/ Biomedical image processing for automatic diagnosis aid	Davide Cavagnino, Maurizio Lucenteforte, Marco Grangetto	The availability of huge amount of digital biomedical imaging data is significantly increasing the possibility to use processing techniques, pattern recognition and computer vision to support medical diagnosis in many fields. The project will develop fundamental research in the field of biomedical image processing in cooperation with medical units of our university and other partners with particular interest in RX, PET images and histological microscopic images.
Watermarking di oggetti digitali multimediali/ Watermarking of digital multimedia objects	Prof. Marco Botta, Davide Cavagnino	Due to the widespread diffusion and distribution of multimedia data like images, videos, sounds and 3D models, there may be the necessity, in some application contexts, to protect the integrity or track the origin of digital objects.
Realtà virtuale, realtà aumentata e la loro convergenza nella mixed reality / Virtual reality, augmented reality and their convergence to mixed reality	Prof. Maurizio Lucenteforte, Davide Cavagnino, Marco Grangetto	A big challenge of the near future will be the convergence of the technologies made available for virtual reality and augmented reality applications. This will permit the definition of a new media, making possible to change radically the way to approach different disciplines, such as medicine, cultural heritage, education, marketing, virtual prototyping and so on.
Biologia dei sistemi: Costruzione ed analisi di modelli/Model Construction and Analysis.Supervisors	Prof. Francesca Cordero, Marco Beccuti	http://compsysbio.di.unito.it/ Description:One challenging aspect of systems biology is that the phenomena under study are not known in details. To have a model that can be analyzed and useful to generate new hypothesis, one starts by describing its structure and then defines the parameters necessary to obtain temporal behaviors of the entities in the model. The project deals to construct the mathematical models that mimic complex biological system. The consistency and correctness of the models constructed within the project will be verified. Finally, the study of the temporal dynamics will be investigate by two following analysis approaches: stochastic and deterministic.