Project Specific Grants / Reserved topic scholarships 2015 - 1st call | Doctoral Program - Information Engineering and Computer Science

Project Specific Grants / Reserved topic scholarships 2015 - 1st call

Area A (Curriculum 1: Computer Science)

IIT / FBK

A1 - Computational Methods for Brain Connectivity Analysis (1 grant)

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: avesani [at] fbk.eu diego.sona [at] iit.it 

FBK

A2 - Flexible machine translation (1 grant) [additional reserved topic scholarship]

Machine Translation (MT) has become reliable enough to concretely support the daily work of professional translators. In computer-assisted translation, machines allow humans to translate faster. In the same framework, human feedback should allow machines to translate better. This objective can be pursued in different ways, for instance by  injecting knowledge about how specific segments (words, phrases, sentences) should be translated. But what about the relation between such segments and the world they refer to and evoke? For instance, what does a human correction of "accept" into "ratify" tell us about the domain and register of the original sentence? And what does it tell us about the proper translation of terms like "home("domicile"), "tell("notify"), "pay" ("indemnify") appearing in the same legal document? To answer these questions, a smart use of human feedback should not only focus on local translation/correction patterns, but also on inference and projection mechanisms to abstract from them. This grant will be offered to students willing to take the challenge by combining research on statistical machine translation, machine learning and distributional semantics in a single "flexible MTframework.
Contact:  negri [at] fbk.eu turchi [at] fbk.eu

A3 - Multimedia information extraction driven by background knowledge (1 grant) [additional reserved topic scholarship]

The objective of the phd, is to develop methodologies and tools for the extraction of facts about events from multimedia document (e.g., commented video, text with pictures, etc.). Differently fron the past we want to apply holistic approach, where where the processing of extract information from the different media are integrated at an early stage, and they are considered as a whole information space, where the interpretation of text affects the interpretation of video and audio and viceversa. Furthermore the information extraction does not happen in the knowledge vacuum, but instead it should exploit existing background knowledge and rationality which is usually expressed in terms of logical theories.
Contact:  serafini [at] fbk.eu

A4 - Vision for Multimedia Understanding (1 grant) [additional reserved topic scholarship]

Multimedia content analysis more and more relies on advanced machine learning to capture the enormous richness of multi-modal sources (commented videos, images with captions, etc.). At the other side, domain specific knowledge is often available to leverage the content analysis task, but effectively encoding it into machine learning (down to the development of task-specific feature representations) is still an open research issue. The goal of this PhD is to progress on the computer vision side of the problem, to go beyond a mono-modal approach where supervisions for learning are provided explicitly. Instead, we will investigate how structured (background knowledge) and semi-structured data (e.g. text captions and descriptions) can be used to provide implicit supervision to enrich the task-specific visual learning capabilities.
Contact: lanz [at] fbk.eu

A5 - Capturing signals from continuous streams of textual data (1 grant) [additional reserved topic scholarship]

Social networks, web sites, blogs and wikis continuously provide a huge amount of signals that can be exploited to respond to business and social needs. Examples of signals are: Is a company growing? Is a vaccine considered safe? Is a politician appreciated? The goal of this thesis is to develop sensors able to capture different types of signals from continuous streams of data and make complex analysis, data comparisons and predictions based on the combination of their output. Sensors will be mainly developed using natural language processing and machine learning techniques. The innovative contribution consists in providing sensors and the applications built on top of them with the ability to interact with the environment they are immersed in to significantly improve their accuracy. Specifically, sensors will actively use social media to collect additional training data to improve the learning strategy, or validate their hypotheses, with particular interest to aspects of active learning, reinforcement learning, and constructive learning.
Contact: giuliano [at] fbk.eu

A6 - Investigating the relationships between spatial structures and urban characteristics (1 grant) [additional reserved topic scholarship]

Nowadays, the majority of people live in cities and the urban areas density is generally increasing. Several works show that the size of cities plays a fundamental role in a systematic acceleration of social and economic life. These gains are applied to different quantities including economic output, wages, patents and epidemics and the average increase in these urban quantities, Y in relation to the population size N, is usually described by a super-linear scale-invariant law. Recently, theoretical works suggest that the origin of this super-linear scaling stems from social interactions. Hence, understanding city’s dynamics and spatial dynamics is of paramount relevance. The recent availability of new large-scale data sets, such as those automatically collected by mobile phone networks and by credit card transactions, open new possibilities of studying city dynamics. The goal of this thesis is to build data-driven models of cities' socioeconomic outcomes using these new sources of data.
Contact:  lepri [at] fbk.eu

Telecom Italia

A7 - Empowering citizens to manage their personal knowledge spaces: personal services based on big personal data analytics (1 grant) [additional reserved topic scholarship]

The project aims at studying analytical methods that enable the understanding of individual behavior and indicators emerging from personal and Big Data. These data are originated from the digital breadcrumbs of human activities and sensed as a by-product of theICT systems used everyday. They record the multiple dimensions of the social life of an individual: payment systems record our purchases; search engines record the logs of our queries; social network services record our connections to other people; wireless networks and mobile devices record the traces of our movements. In this context, it is becoming more and more important to study methodologies for integrating different kinds of data (GPS data, phone data, purchasing data, social data, and so on) and to explore the human digital multidimensionality. These methodologies pave the way for the development of advanced behavior prediction models and personalized recommendation systems, enabling the design of added value personal application based on the individual behavior. Clearly, this scenario can introduce new challenges and new open doors with respect to the individual privacy issues, that require the evaluation of privacy risks deriving from the detailed inferences which are possible on the individual personal sphere.
Contact:  andrea.passerini [at] unitn.it michele.vescovi [at] telecomitalia.it

 

Area B (Curriculum 1: Computer Science)

Department of Information Engineering and Computer Science

B1 - Guidance solutions for robotic assisted devices (1 grant)

The research aims to develop and test solutions for guiding a senior user through the execution of a motion plan using a robotic assistant. The PhD candidate will develop solutions based on electro-actuated motors and turning wheels. The algorithms developed will be clinically tested in a real environment to verify teh practical applicatbility of the robotic walker as a rehabilitation tool.
Contact: luigi.palopoli [at] unitn.it

B2 - Motion planning algorithms for assisted robotic devices (1 grant)

This research will be on finding solutions for motion planning of a robotic walker across a crowded environment. The PhD candidate will investigate on the following themes:
1. dynamic models for describing cooperation and competitions between humans in crowded spaces
2. logic formalism to describe the user's requirements and golas
3. lightweight algorithms for synthesis of socially aware optimal plans.
Contact: luigi.palopoli [at] unitn.it

B4 - Algorithms for planning and control of mobile robots and drones (1 grant) [additional reserved topic scholarship]

Th PhD student will work on a selection of topics cutting across two different disciplines: motion planning and control. Considering different types of vehicles with their motion constrained by the environment, the student will have to identify sequence of possible actions that determine units of motion compatible with different types of constraints (both geometric and of different type) and associate a control law that implement each of them. Then she/he will seek appropriate ways to combine such elementary actions into appropriate sequences that produce a system--level expected effect.
The first year will be devoted to studying the state of the art, then the student will prepare a detailed thesis proposal that will have to be implemented into software design tool and experienced on such phyisical systems as robotic walkers, flying drones and car--like robots.

Contact: luigi.palopoli [at] unitn.it

FBK

B3 - Designing incentive models for citizen cooperation in smart communities (1 grant) [additional reserved topic scholarship]

In smart communities, citizens contribute actively to the adoption of IT solutions to improve the quality of life in their city or region. Incentives are recognized as key enablers for the construction of smart communities, both to engage citizens and to promote their active participation. The objective of this research activity is to design and experiment new incentive models that are effective in engaging citizens. In particular we will focus on *data-driven* incentives that are able to promote *cooperative* behaviors among citizens.
Contact:  lepri [at] fbk.eu pistore [at] fbk.eu

 

Area C (Curriculum 1: Computer Science)

Department of Information Engineering and Computer Science

C1 - Advancing active and healthy ageing with ICT (1 grant)

The PhD candidate will work in the EU project ACANTO: A Cyberphysical social Network using robot friends, within the general theme of advancing active and healthy ageing with ICT, specifically with service robotics within assisted living environments. The present research theme will focus on the definition of appropriate recommendation of the most suitable social activities personalized to individual users or elder adults communities profiles in a given intelligent environment.
Contact: maurizio.marchese [at] unitn.it

C7 - Software evolution through release planning (1 grant) [additional reserved topic scholarship]

TFormally define and solve the requirements planning problem, evaluate the prposed framework and tools with case studies adopted from software practice.
Contact: jm [at] cs.toronto.edu

C8 - Requirements patterns and anti-patterns (1 grant) [additional reserved topic scholarship]

THow to discover patterns and anti-patterns and how to use them during requirements engineering.
Contact: jm [at] cs.toronto.edu

FBK

C2 - From Laws to their Implementation into Socio Technical Systems Regulations and Policies (1 grant) [additional reserved topic scholarship]

This proposal intends to investigate the connection between laws, in particular privacy related laws, and the policies implemented in an organisation and enforced via information systems. We envisage an approach that considers the modelling of laws as preliminary step for the implementation of reasoning techniques to check the alignement of the policies with respect to the laws, maintain traceability between the two artifacts and search for possible alternative law implementations into policies.
Contact:  ranise [at] fbk.eu susi [at] fbk.eu

C3 - Safety and Dependability Assessment with Formal Methods (1 grant) [additional reserved topic scholarship]

The design of complex critical systems requires the ability to analyse the behavior in presence of faults, and to produce artifacts such as Fault-Trees and FMEA tables. Traditional techniques, applied in various sectors such as railways, avionics and space, are mostly manual, and are thus error prone and very costly. Recently, the use of formal methods has been proposed for the automated analysis of safety and dependability. Scalability and usabilty, however, remain important issues to be addressed. The thesis concerns the investigation of novel techniques for safety and dependability, able to automatically construct artifacts closer to the ones required in practice, and to increase the scalability by means of compositional reasoning. The aim is to obtain a formal environment for safety and dependability able to deal efficiently with large system-level designs taking into account also the interaction between discrete and continuous dynamics as required by industrial practical designs.
Contact:  cimatti [at] fbk.eu

C4 - Cyber Physical Systems Design and Verification (1 grant) [additional reserved topic scholarship]

Cyber-physical systems (CPSs) are systems of collaborating computational elements controlling physical entities and connected to the internet. CPSs have applications in several areas such as aerospace, automotive, chemical processes, civil infrastructure, energy, healthcare, manufacturing, transportation, entertainment, and consumer appliances. CPSs are required to carry out critical functions, and must implement an ever increasing number of complex functions. The development of predictable CPSs is thus a fundamental challenge, and requires the definition of model-based formal techniques and tools able to analyze their behaviors taking into account both the discrete and continuous dynamics, and tools to support the analysis of the design space exploiting the structure of the system compositionally. The thesis will investigate the relation (and will bridge the gap) between currently available techniques for CPS modeling, and formal verification techniques for continuous/discrete systems taking into account compositional issues to enable for the verification of such complex heterogeneous designs.
Contact:  cimatti [at] fbk.eu

C5 - Adopting Human Language Technologies to facilitate the access to on-line services (1 grant) [additional reserved topic scholarship]

Public administration is relying in an increasing way on on-line interactions to deliver services to citizens. An important obstacle to the adoption of these on-line services by citizens is that they are often accompanied by descriptions and explanations  written in technical and legal language, and hence difficult to understand for non-experts. In this research, we intend to adopt Human Language Technologies to adapt these textual descriptions, annotate and personalize them to the citizen interacting with the on-line service. In particular, this approach will include language simplification of administrative documents at different levels (lexical, syntactic and semantic) as well as semantic enrichment through external knowledge sources leveraged from citizens, civil servants and different information repositories. From a research point of view, the novelty of this approach will lie in the focus on the user characteristics and on the choice of different simplification strategies tailored to users’ profiles.
Contact:  satonelli [at] fbk.eu pistore [at] fbk.eu

C6 -  Combining planning and machine learning techniques for the re-planning of clinical pathways (1 grant) [additional reserved topic scholarship]

In the last decades, the use of IT systems for supporting business activities has notably increased, thus opening to the possibility of monitoring business processes and performing on top of them a number of useful analysis. This has brought to a large diffusion of tools that offer business analysts the possibility to observe the current process execution, identify deviations from the model, perform individual and aggregated analysis on current and past executions, thus supporting process model re-design and improvement. Unfortunately, a number of difficulties may arise when exploiting information system data for monitoring and analysis purposes. Among these, on-line monitoring of data may identify deviations from the model that need to be handled and solved while the process is running so that the process execution can continue in an acceptable way, ideally by returning to a compliant execution as fast as possible. An example of this scenario happens in the medical domain, where clinical paths for certain conditions (e.g., a pregnancy) can deviate from the suggested medical path due to e.g., unavailability of facilities or exceptions not captured in the model, but need to go back and follow the "normal path" after the local deviation is performed. The aim of this thesis is investigating how to exploit, adapt and combine techniques and approaches borrowed from different research fields, in particular planning and machine learning, to advance the existing services for process (re-)planning. Planning techniques will make usage of the procedural knowledge encoded in the medical guidelines (following the approach of e.g., [1]), while machine learning approaches will exploit the presence of monitored similar data (starting from frameworks such as [2]).  To this purpose, several are the challenges to be faced in the work as, for example, (i) the capability to represent and reason about secondary aspects for business processes such as data, time, resources; (ii) the capability to align execution information with models; (iii) the capability to realize the abovementioned analyses at run-time. The work will put together theoretical and methodological aspects, including for example the problem conceptualization and representation, as well as implementation and optimization ones, aimed at the development of process analysis services and tools.
Contact:  ghidini [at] fbk.eu cleccher [at] fbk.eu

Telecom Italia

C9 - User's drivers investigation and collaborative design methodologies  for personal data driven services (1 grant) [additional reserved topic scholarship]

Quantify yourself – The objective of this PhD grant is investigating the needs, the motivations and the user drivers toward their acceptance of the monitoring and tracking of their personal life through the collection and exploitation of personal data. In particular this PhD aims at studying the pertinence and acceptance of tools for self-reflection in many different contexts (for instance family life, shopping behaviors, individual tasks and aspects of the daily life, etc.). The work include empirical investigations with the goal of designing and developing of technologies and real services in a personal data driven context.
Contact:  zancana [at] fbk.eu michele.caraviello [at] telecomitalia.it

C10 - Emotion recognition from wearable physiological-sensors to dynamically assess media and social content (1 grant) [additional reserved topic scholarship]

In the last years wearable (sensors, such as wristbands, clocks, etc.) shown among the technologies which have experienced the higher increase of attention and expectations, both in terms of applicative domains and impact on user life. Moreover they witnessed a fast and significant growth under many perspective: in terms of production (variety of producers, sensors, features and associated algorithms), in terms of adoption and diffusion, in terms of quality and potentials improvements.
These technologies allow (in a non-invasive way) the continuous collection of physiological data from the users, in order to support them in their daily activity and to provide personal awareness, e.g. in quantify-self scenarios. The aim of this PhD thesis is to consolidate the know-how and the results reached in previous studies (performed in cognitive science with precise, medical-quality sensors) transferring them to off-the-shelf wearable sensors, with the specific focus of emotion-detection scenarios in the media domain, such as for the evaluation of content (e.g. images, videos,...), or in social applications such as social-networking, social-sharing and comparison.

Contact:  niculae.sebe [at] unitn.it michele.vescovi [at] telecomitalia.it

C11 - Real-time Anomaly Detecion on Heterogenous Data Streams (1 grant) [additional reserved topic scholarship]

Anomaly detection is a critical step in order to get a better understanding of complex systems and gather insights about them. The topic becomes even more interesting and challenging when the anomaly has to be detected in real-time on data streams that have different provenance. The state of the art is rich of effective solutions applicable when data is homogenous and "at-rest", but it lacks solutions for heterogeneous data "in-motion".
The goal of this PhD thesis is the development of advanced data analytics techniques that exploit modern advances in logical data warehousing and parallel query execution/optimization in order to enable the analysis of Big Data streams. Using the 4Vs of Big Data, this PhD thesis aims at addressing Velocity and Variety at the same time.
In this PhD the candidate will have the possibility to apply his/her findings on real data streams like the records generated from the mobile phone activity of the TIM users (SMS, call and internet traffic) and social network data.

Contact:  velgias [at] unitn.it roberto.larcher [at] telecomitalia.it

 

Area D (Curriculum 2: Telecommunications)

Department of Information Engineering and Computer Science

D1 - Automatic techniques for parameter estimation from satellite remote sensing data (1 grant)

This PhD Position requires specific know-how and basic knowledge in the field of remote sensing. The research activity will be focused on the development of automatic techniques for analysis and parameter estimation from remote sensing data. The work will be focused on either information extraction from radar sounder data (in particular in the framework of the "Radar for Icy Moon Exploration (RIME)" instrument on board of European Space Agency "JUpiter ICy moon Explorer (JUICE)" or on paramete estimation from Earth Observation data. and The research will be carried out at the Remote Sensing Laboratory (RSLab) in the Department of Information Engineering and Computer Science of the University of Trento (see http://rslab.disi.unitn.it).
Contact: lorenzo.bruzzone [at] unitn.it

D2 - Techniques for the simulation and the automatic analysis of data acquired by the Radar for Icy Moon Exploration (RIME) (1 grant)

The activity will be carried out in the framework of the "Radar for Icy Moon Exploration (RIME)" instrument on board of European Space Agency "JUpiter ICy moon Explorer (JUICE)". RIME is a low frequency radar sounder (i.e. a ground penetrating radar operated from a satellite platform) designed to investigate the Jupiter Icy Moons (i.e. Ganymede, Europa, and Callisto). The activity will be related to the simulation of the performance of the radar sounder and to the definition of automatic techniques for the analysis of the radargrams acquired by the sounder for the definition of the ground segment processing chain. The research will be carried out at the Remote Sensing Laboratory (RSLab) in the Department of Information Engineering and Computer Science of the University of Trento (see http://rslab.disi.unitn.it).
Contact: lorenzo.bruzzone [at] unitn.it

FBK

D3 - Automated Security Analysis of Cloud and Mobile Applications (1 grant) [additional reserved topic scholarship]

With the convergence of the social, cloud, and mobile paradigms, information and communication technologies are a ffecting our everyday personal and working live to unprecedented depth and scale. We routinely use online services that stem from the fruitful combination of mobile applications, web applications, cloud services, and/or social networks. Sensitive data handled by these services often flows across organizational boundaries and both the privacy of the users and the assets of organizations are often at risk.
Solutions (e.g., security protocols and policies) that aim to securely combine the ever-growing ecosystem of online services are already available. But they are notoriously diffi cult to get right. Many security-critical applications and services have been designed and developed only to be found flawed years after their deployment. These flaws are usually due to the complex and unexpected interactions of the protocols, policies, and services together with malicious agents. Since these weaknesses are very di fficult to spot by traditional verifi cation techniques (e.g., manual inspection and testing), security-critical applications are a natural target for formal method techniques.
In this thesis, we intend to investigate how automated formal method techniques can be adapted and combined to spot vulnerabilities or to certify their absence.
Contact:  ranise [at] fbk.eu

EURAC - European Academy of Bozen/Bolzano

D4 -Developing innovative algorithms for the use of Sentinel data to infer biophysical properties and their dynamics in the Alpine region (1 grant) [additional reserved topic scholarship]

The use of remote sensing for monitoring the environment is a widespread technique. In this context the launch of the Sentinel satellites is starting a new era for what concern temporal and spatial resolution. This new generation of Earth observation (EO) data will be new ground for improving the retrieval of biophysical parameters due to their new improved performances. Especially mountain areas requires specific development due to their peculiarities such as topography and landscape heterogeneity. Amont the most interesting parameters, it is worthwhile mentioning soil moisture, snow cover/snow status, leaf area index. This PhD project will therefore focus on developing and validating innovative algorithms by using Sentinel data (mainly Sentinel 1 and 2) specifically for mountain areas and generating time series of the main biophysical parameters over the Alpine areas.
Tasks:

  • Making use of multiple satellite sensors (i.e., Sentinel 1, Sentinel 2) for retrieval of main biophysical parameters;
  • Developing innovative algorithms for inferring biophysical parameters from optical and radar images;
  • Making use of ground data and proximal sensing for the product validation;
  • Taking part in field campaigns;
  • Publishing in scientific journals.

The research will be carried out at the Institute for Applied Remote Sensing (www.eurac.edu).
Contact: Paola Winkler, tel. +39 0471 055 970

D5 - Integrating interferometric SAR techniques with ground data to monitor terrain movement in Alpine area (1 grant) [additional reserved topic scholarship]

The EURAC Institute for Applied Remote Sensing develops methods, tools, products and services for the Alps and other mountainous areas, addressing current environmental problems in the context of global change. For these tasks the use of remote sensing for monitoring the environment is a widespread technique. The recent launch of the Sentinel satellites has started a new era with regards to temporal and spatial resolution.The PhD project will combine the activities related to landslide and ground deformation monitoring, using in-situ-measurements and remote sensing data such as satellite-based InSAR. The PhD candidate will apply state-of-the art interferometric techniques and develop methods to integrate this information with other data such as GPS information.
Tasks:

  • Applying InSAR techniques (including multi-interferometry) to COSMO-SkyMed, Sentinel 1 images and other available SAR data
  • Analysis and interpretation of landslide / ground movement monitoring data (e.g. GPS, satellite-based InSAR)
  • Developing techniques to integrate satellite and ground data
  • Carrying out field measurements in alpine terrain (e.g. using differential GPS)
  • Publishing in scientific journals and presentation of research results at conferences.

The research will be carried out at the Institute for Applied Remote Sensing (www.eurac.edu).
Contact: Paola Winkler, tel. +39 0471 055 970