You are here

Phd Thesis Proposals

The following PhD positions are currently available within the DKM unit:

Bayesian reasoning for statistical relational learning (2 positions)

Current approaches in statistical relational learning are based on undirected graphical models such as Markov Logic Networks. State of the art algorithms for statistical inference cover the Maximum Likelihood (ML) and Maximum a Posteriori (MAP) tasks, but not so much attention has been devoted to Bayesian Inference. Due to the high complexity of the models that can be generated, statistical inference is approximated using sampling methods. Recently, we proposed a study about Bayesian Inference in hybrid graphical models (i.e., models composed of discrete and continuous random variables); the advantage of Bayesian inference is that, it’s a truly statistical inference and it is very robust to overfitting training data. We design a variational method to solve the “exact inference”. However, to perform Bayesian inference, combinatorial problems on the discrete variables must be solved in a more efficient way, and this is still an open problem. The objective of this thesis, is to extend such proposals and to make scalable.

CONTACT and INFO: Luciano Serafini

SCHOOL: University of Trento - Doctoral Program in Information and Communication Technology

DEADLINE: May 20, 2019, hrs. 04.00 PM (Italian time, GMT +2)

NOTE: In the application specify that you are interested in the Project Specific Grants (reserved topic scholarships)  A7 - Bayesian reasoning for statistical relational learning

APPLICATION SITE: https://ict.unitn.it/education/admission/call-for-application

Incremental learning of abstract planning models via acting in a real environment (2 positions)

Autonomous agents, such as robots, chat-bots, self-driving cars, soft-bots etc., need to plan their actions in order to achieve their goals. For this reason, they should know the environment in which they operate and the effects of their actions on the environment. These information are usually encoded in the so-called “planning domain”, which, need to be “programmed off line” when the agent is programmed. However, the environment is dynamic and can have unpredicted changes; therefore, the agent should be able to adapt to unexpected situations. Furthermore, the effect of actions could be vary complex and unknown since the beginning; the agents should be able to learn action effects while acting. The objective of the Ph.D is to develop the necessary theory and the algorithms that allow an agent to incrementally learn a discrete, compact, and semantically rich representation of the planning domain in an environment in which it is supposed to interact. This representation is formulated in a form of a discrete planning domain.
CONTACT AND INFO: Luciano Serafini and Paolo Traverso

POSITION 1 at University of Trento - Doctoral Program in Information and Communication Technology

  • DEADLINE: May 20, 2019, hrs. 04.00 PM (Italian time, GMT +2)

POSITION 2 at University of Padova - PhD Course: BRAIN, MIND AND COMPUTER SCIENCE in agreement with Fondazione Bruno Kessler

  • DEADLINE: 14 MAY 2019, 13:00 CEST

Default in contextualized knowledge representation

Contextual representation of knowledge has been one of the main approaches that allow to represent large knowledge bases in which knowledge is relativized w.r.t. the context (situation, set of hypothesis) in which it is supposed to hold. Default reasoning in contextual knowledge is a new research issues and play an importan role in default knowledge propoagation. In this thesis we will investigate on extending the methodologies for default contextstual knowledge propoagation.

CONTACT and INFO: Luciano Serafini

SCHOOL: University of Trento - Doctoral Program in Information and Communication Technology

DEADLINE: May 20, 2019, hrs. 04.00 PM (Italian time, GMT+2)

NOTE: In the application specify that you are interested in the Project Specific Grants (reserved topic scholarships)  C3 - Default in contextualized knowledge representation

APPLICATION SITE: https://ict.unitn.it/education/admission/call-for-application

Understanding multimedia with the help of background knowledge

This phd has the objective of extracting events from commented videos exploiting background knowledge available in the semantic web. This phd should develop a holistic approach, where the process of extracting information from the video, and from the associated text are integrated and can affect each other at any stage. This implies that video stream and textual stream are considered as a whole information space and their interpretations are not independent. Furthermore, video-text interpretation should not happen in the knowledge vacuum, but it should exploit the existing large amount of background knowledge available in the semantic web under the form of ontologies and RDF data. Nowadays--in contrast with the early years of AI when knowledge acquisition was a bottleneck--large amount of commonsense knowledge is available in the semantic web, but it cannot be easily exploited by the state-of-the-art approaches to video and text analisys. The thesis should investigate on how to extend and adapt algorithms for video and text analysis in order to inject background knowledge. The thesis, to reach it's objective, should combine techniques in machine learning--for processing low level data--with automated reasoning--to manage with high level semantic knowledge.

CONTACT and INFO: Luciano Serafini

SCHOOL: University of Padova - PhD Course: BRAIN, MIND AND COMPUTER SCIENCE in agreement with Fondazione Bruno Kessler

DEADLINE: 14 MAY 2019, 13:00 CEST

NOTE: In the application specify the priority on the Scholarship - Understanding multimedia with the help of background knowledge(see figure 7 of the document describing application instructions)

APPLICATION SITE: http://hit.psy.unipd.it/BMCS/admission