2 PhD Studentships in collaboration with the University of Trento
As part of the FBK International PhD Program, the HLT-NLP research group at Fondazione Bruno Kessler han an open call for 2 PhD studentships in collaboration with the University of Trento (ICT International Doctoral School) on the following topics:
- Cross-Language Semantic Parsing (1 grant)
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
- Flexible Dialogue Models for Conversational Agents and ChatBots (1 grant)
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.
Closing Date: August 31st, 2017
The selected candidates will join the HLT-NLP Group at Fondazione Bruno Kessler, in Trento (Italy).
Detailed instructions and further details on the PhD call are available here.
Contact: Bernardo Magnini