Page related to first call year 2017 available at http://ict.unitn.it/application/project_specific_grants_2017_1st_call
COSBI - The Microsoft Research – University of Trento Centre for Computational and Systems Biology
The rapidly growing body of biomedical knowledge in scientific publications calls for the development of new approaches to knowledge extraction and discovery from unstructured texts. Specifically the candidate is expected to be confident in the biological domain and will focus on the design and development of novel natural language processing methods in this area exploiting the power of deep learning and machine learning algorithms, in order to learn and abstract relations and facts in a wide range of biological applications (eg. biomolecular pathways discovery, reaction rates extraction, meta-analysis from multiple sources statistical data, visualization tools for easier biomedical result interpretation, etc.).
DISI - Department of Information Engineering and Computer Science
The evolution of urban spaces toward a smart living environment requires a continuous evolution of the communication networking environment. A smart city, indeed, is a place where the overwhelming amount of information that is characteristic of our society is correctly managed by the People for the People. This little, short sentence hides a revolution in the attitude, and a scientific and technological leap from the current situation of centralized communications and data handling and management by incumbent operators (from telcos to information arbitrators like Facebook, Google or Amazon) to a distributed ownership and management of networks, to a public (by the People) control of data usage. This PhD program focuses on novel network layer protocols, novel network management, novel applications that can move our living spaces toward the idea of a Smart City. Topics include, but are not limited to, routing, monitoring, disruption recovery, mobile and vehicular, including communications with Vulnerable Road Users (VRUs), beyond 5G access, and resource usage and allocation.
Rapid technological developments challenge the personal data protection. Technologies allow for collecting and sharing personal and sensitive data to pursue activities in order to enhance quality of life and create new business opportunities. The main research problem to be addressed during the PhD is the development of a conceptual framework for the representation of privacy concerns in large and complex socio-technical systems and the automated verification of privacy policies coming from notational and international regulations (such as the new GDPR). We expect to develop a novel privacy-oriented modelling language, a systematic methodology to follow the principle of privacy-by-design, reasoning techniques for the automated verification of privacy policies over privacy models, and a prototype tool to support the entire framework. Finally, results of the research will be validated by means of real cases from research and industrial projects.
The research activity is focused in the field of satellite remote sensing systems and is devoted to the development of advanced automatic methods for the analysis and the fusion of multisource remote sensing images and data. The goal is to define, design, implement and validate methods based on pattern recognition and machine learning for the analysis of remote sensing images acquired by different sensors and for the integration/fusion of these images with other ancillary data. Different paradigms will be studied related to the most recent methodological developments in the framework of information extraction and fusion. Specific attention will be devoted to the integration of remote sensing data with the data derived through physical-based models.
The general methods developed will be then customized on the problems of cryosphere monitoring and hydrological parameter estimation. In this framework, an extensive validation of the developed methods will be carried out by using different kinds of remote sensing data, physical-based models and ancillary data. The PhD student will work at the Remote Sensing Laboratory of the University of Trento.
For more information on the activity of RSLab refer to: http://rslab.disi.unitn.it/ or contact Prof. Lorenzo Bruzzone firstname.lastname@example.org
The research activity that will be developed is related to the definition and the design of a new Earth Observation mission based on a radar sounder for the analysis of the Earth subsurface. The main goals of the mission are related to the study of the sub-surface of the polar ice sheets and of the subsurface of arid areas. This has a huge impact on the study of the cryosphere and the climate, as well as on the detection and analysis of the water table in the desert. The research activities related to the PhD position can address different specific directions (to be agreed with the selected candidate) related to the main problems to be solved for this kind of mission, including the definition, design, implementation and validation of: 1) data simulation techniques for radar performance assessment; 2) strategies and techniques for mitigating the effects of the ionosphere on the radar signal; 3) radar signal processing techniques; 4) techniques for clutter reduction; and 5) data analysis techniques for the automatic extraction of information from radargrams for the generation of mission products. The PhD student will work at the Remote Sensing Laboratory of the University of Trento.
For more information on the activity of RSLab refer to: http://rslab.disi.unitn.it/ or contact Prof. Lorenzo Bruzzone email@example.com
EIT Digital, Centro Ricerche Fiat S.c.p.a. and Department of Industrial Engineering
A7 - Artificial Intelligence (AI) solutions for Autonomous Driving enacted with Deep Neural Network technologies and humanlike information processing architectures (1 grant) [additional reserved topic scholarship]
The proposed PhD program will addresses a number of relevant research challenges for the automotive industry, with particular focus on autonomous driving, such as:
- fusion of heterogeneous data coming from various on-board sensors;
- mimicking loops in the brain in relation to action selection;
- predicting expected driver perception;
- leveraging on Deep Neural Networks (DNN) and building on the NVIDIA Software Development Kit
The scholarship is subject to the acceptance of nondisclosure agreements and the assignment of the outcomes, according to the rules of the project H2020 Dreams4Cars
EIT Digital, Engineering Ingegneria Informatica SpA and Fondazione Bruno Kessler (FBK)
This thesis will leverage on the City Enabler concept, namely an IT solution that offers a collaborative plaza allowing all relevant private/public stakeholders to work together, both for publishing data and for exploiting them to create novel urban services. In particular, the thesis will investigate innovative approaches, techniques and tools for the data and service management, specifically for the “intelligent” retrieval and presentation of information distilled from PA (Public Administration) documents.
The thesis will lead to the development of an information retrieval tool, allowing flexible, multi-layered search of documents content and ranking the search output according to stakeholders’ preferences. The results will be evaluated in real-world pilots in the scope of research and industrial projects in the Autonomous Province of Trento.
This thesis will concentrate on developing a data-driven computational framework to predict and enable the so-called urban vitality. Urban vitality is referred to as the synergy arising from a variety of commercial and entertainment opportunities, and a dense socially heterogeneous pedestrian population. More precisely, the goal of the thesis is to: i) identify the urban conditions (e.g. land use mix, mobility, safety perception) that magnify and influence urban life; ii) study their relationship with cities’ outcomes like innovation, vitality, quality of life, and social cohesion, etc.; and iii) envision data-driven guidelines and tools (dashboards, simulation tools) responsive to the real-time demands of citizens and policy-makers.
EIT Digital, Centro Ricerche Fiat S.c.p.a. and Fondazione Bruno Kessler (FBK)
Intelligent Transport Systems (ITS) applications are supported by several communication technologies, each one with its frequency range and specific features. Evaluating the performance of different network options for V2X communication that ensure optimal utilization of resources is a prerequisite when designing and developing robust wireless networks for ITS applications. 5G networks are expected to leverage on virtualization of network resources in order to serve over the same infrastructures applications and services characterized by highly heterogeneous requirements, the so called verticals. The thesis will investigate the potentialities introduced by the 5G network for the automotive domain, identifying use cases and scenarios, and deriving requirements for the M(V)NO. The identified solution will be experimentally validated in a lab environment (Hardware in the loop or simulated scenarios) and in a more realistic conditions.
The scholarship is subject to the acceptance of nondisclosure agreements and the assignment of the outcomes.
EIT Digital and TIM
The aim of this research activity is to advance in the application of machine learning and recommendation system techniques, for developing a new class of recommendation systems addressing the personalized construction of product/service bundle offers.
The devised solution will advance the current state-of-the-art, being able to tackle the problem of building up (rather than choosing) the recommended “objects”, assembling them from building blocks. The system will incorporate trends information and revenue management aspects in order to maximize the benefits for the provider, thus sustaining up-selling/cross-selling, increasing revenues, as well as for potential customers, in terms of their overall satisfaction and price.
Fondazione Bruno Kessler (FBK)
Semantic Parsing is meant to provide a formal representation of the meaning of a sentence, with the aim of capturing “who is doing what to whom”, which is expressed in natural language texts. As far as applications are concerned, Semantic Parsing is a crucial component, among the others, for question answering, conversational agents, information extraction and aspect-based sentiment analysis.
Most current technologies in Semantic Parsing (e.g. semantic role labeling) are based on supervised machine learning and need large amounts of manually annotated data. The goal of this PhD Thesis is to develop light supervision approaches able to improve portability both among languages and application domains. The candidate is expected to investigate innovative research directions, taking advantage both of existing lexical resources (e.g. PropBank), entity linking technologies, and semantic projections through cross-language alignment.
The PhD candidate will be responsible for the development and application of deep learning methods and techniques to leverage the information contained in the Electronic Health Records (EHR). The objective of the work will be to establish disease models and patient representation using large set of data from EHR, Personal Health Records (PHR) and other sources of data (such as quantified self) to predict the risk of chronic diseases both in patients and healthy population. The ideal candidate will have a background in machine learning, with experience in deep learning and the associated frameworks (such as TensorFlow, Theano or Keras) applied to high-dimensional, irregular, temporal and sparse data. Knowledge of RNNs would be desirable. The post offers an exciting opportunity to work at the cutting-edge of machine learning in medicine, using real patient data to model and predict onset of chronic diseases, with the potential to impact delivery of care paradigms.
Informal enquires about this post can be sent to Dr. Venet Osmani (firstname.lastname@example.org).
Conversational agents are designed to interact with users in multiple domains on several topics using natural language. Many chatbots have been deployed on the Internet (social media, e-commerce websites, just to mention a few) for the purpose of seeking information, question answering, coaching tasks, online shopping, and so on. Usually these applications work in a strictly limited domain with a clear and well defined dialogue structure, with little adaptation capabilities to the contextual and social situation.
The goal of this PhD Thesis is to improve the portability of dialogue modeling both among languages and application domains. We aim at extracting dialogue schemas from diverse sources (e.g. social networks) and to develop new methodologies to combine these schemas. The objective is to allow for a greater dialogue flexibility and re-planning capabilities when the conversational agent is faced with unknown or unexpected situations.
The candidate is expected to investigate innovative research directions in dialogue modeling, integrating advanced machine learning technologies (e.g. deep learning) and knowledge based technologies.
This PhD is based on a collaboration with Adeptmind inc, and the candidate will have the opportunity to spend some periods in the company.
The emerging paradigms for cloud computing involve multi-level systems where there exist different owners, multiple islands of the infrastructure and possibly edge components integrating additional hardware at the edge of the system. Hence, the decision on how to allocate resources in such a dynamic and complex system involves the decision of several actors having possibly conflicting objectives, and the resulting operating point of the system is decided by their utilities, the economic figures into play and the mechanisms implementing pricing and resources matching. The candidate will perform theoretical, algorithmic and simulation research on resource allocation problems in this context, using mathematical tools from game theory, optimal control and algorithmic. Master degree in Mathematics, Physics or Computer Science is preferred. The PhD grant is jointly supported by Fondazione Bruno Kessler (FBK), for further information please refer to Francesco De Pellegrini.
Videos are key elements for gathering evidence of an event occurring in the scene. In security and forensics, surveillance systems are widely employed to control indoor and outdoor environments, in order to prevent crimes. In this context, low-power imaging is an emerging technology, allowing vision systems to operate for months, powered with batteries, with no need for infrastructures.
The proposed research activity aims at developing novel CMOS vision sensor architectures integrating image sensing with efficient real-time processing on the same chip in order to detect anomalous events in the scene and taking low-level decisions.
Blockchain can offer IoT devices a playground where they can be identified without the need of involving central trusted authorities (decentralised identity control) and the possibility to operate and interact within a trust-less environment. One of the big challenges of blockchain technologies is the lack of privacy mechanisms that avoid users to openly and publicly publish personal data on the blockchain public ledger.
Objective of this PhD is the study on novel processes, paradigms methods to preserve privacy when discovering IoT devices, when authenticating and offering access to them through blockchain technologies.
This PhD thesis will investigate approaches, techniques and tools to support the public administrations in the process of defining, publishing and managing open data and open services. In particular, the main theme of the thesis will be a methodology for selecting the data to open, offer and enrich it as open services guided by the need to address the requirements of various stakeholders in a city (groups of citizens, associations, companies, politicians...).
The methodology will then be implemented through specific web tools. The tools will in turn help the data provider to monitor the quality of the data generated, measure the impact of the applications created and evaluate the development of new citizen-oriented services.
We will also investigare visual languages that make possible also for non-skilled users some basic tasks of data visualization.
The results will then be evaluated in real-world pilots in the scope of research and industrial projects in the city of Trento.
[description updated - August 1, 2017]
OpenStreetMap is the world's largest collaborative collection of geo-referenced data useful for creating maps (and not only).
This crowdsourcing project is in constantly growing (there are over 4,000,000 registered users at the beginning of 2017) that produces data with lots of detail.
The maps of OpenStreetMap and the data produced are now used by so many public administrations and companies and with a great success in humanitarian response and economic development.
However, one of the most commonly asked questions is the one related to the reliability of the data.
In these years the research is in particular addressed to the comparison with public administration data or in the study of the community.
The aim of the phd thesi is to identify a methodology that validates and predicts the quality of openstreetmap data on the basis of comparison with data from territorial agencies and also on the quality of the users participating in the project.
Contact: email@example.com - firstname.lastname@example.org
The Remote Sensing for Digital Earth Unit at Fondazione Bruno Kessler is developing research activities in the context of sub-surface modelling and understanding by information extraction from radar sounder data. Satellite-borne radar sounders have been used to measure sub-surface properties of the Moon and Mars, whereas airborne instruments have been operated for Earth observation. Nowadays activities are ongoing on the design of satellite borne radar sounders for the study of Jupiter Icy Moons and Earth. In this context, the PhD research activity will be focused on the design and development of radar signal processing methods, automatic algorithms for information extraction and tools for product generation from radargrams.
Fondazione Bruno Kessler (FBK) and Istituto Italiano di Tecnologia (IIT)
The PhD aims at carrying out research activity on machine learning methodologies for brain connectivity data analysis. The main goal is to design and to deploy machine learning algorithms for open challenges such as the detection of the main structural and functional pathways of the brain, the characterisation of the differences with respect to altered brain connections, the inter-individuals analysis of brain connectivity structures.
The PhD grant is jointly supported by Fondazione Bruno Kessler (FBK) and Istituto Italiano di Tecnologia (IIT). The research activity will take place at the Neuroinformatics Laboratory (NILab) and Pattern Analysis and Computer Vision Laboratory (PAVIS).
Contact: email@example.com - firstname.lastname@example.org