UCL  IRIS
Institutional Research Information Service
UCL Logo
Please report any queries concerning the funding data grouped in the sections named "Externally Awarded" or "Internally Disbursed" (shown on the profile page) to your Research Finance Administrator. Your can find your Research Finance Administrator at https://www.ucl.ac.uk/finance/research/rs-contacts.php by entering your department
Please report any queries concerning the student data shown on the profile page to:

Email: portico-services@ucl.ac.uk

Help Desk: http://www.ucl.ac.uk/ras/portico/helpdesk
 More search options
Prof Sebastian Riedel
Appointment
  • Professor of Natural Language Processing & Machine Learning
  • Dept of Computer Science
  • Faculty of Engineering Science
Biography

After having received a Dipl. Ing in Computer Science and Engineering at the Technical University Hamburg-Harburg in 2003, I went on to do an MSc (2004) and PhD (2008) in the University of Edinburgh. In 2008 I worked as researcher at University of Tokyo and the Database Center for Life Sciences. After being postdoc and research scientist at UMass Amherst I started as a lecturer at UCL in 2012.

Research Groups
Research Themes
Research Summary

Whenever we advance our understanding of the world, we write down our findings in publications, patents, webpages or the like. This results in a vast and ever-increasing body of text, impossible to effectively access and comprehend. The overarching goal of my research is to extract and distill this knowledge. 

The cornerstone of my research are probabilistic models that capture global dependencies between data and meaning. I have used such models to develop several state-of-the-art information extractors. The basis of this success has been my broader research focus on four basic questions:

Models: What global dependencies should be captured? One of my core hypotheses is that different levels of linguistic annotation should be tightly connected, allowing high level semantic information to inform low-level processors.

Inference: How can we efficiently reason in the presence of such dependencies? I believe that one answer to this question is mathematical optimization and the triad: decomposition, relaxation and exploitation of symmetries.

Learning: How can we learn the parameters of such models with minimal supervision? Often knowledge of the type we like to extract exists already in structured form. By aligning this knowledge with text, we can train extractors without having to manually annotate documents.

Representation: How do we represent the extracted knowledge? I have developed methods that can extend relational schemas by reading text. My current hypothesis is that text itself should be the interface to knowledge, and latent representations of meaning should only means to the end of better predicting “what could be said." 




Please report any queries concerning the data shown on this page to https://www.ucl.ac.uk/hr/helpdesk/helpdesk_web_form.php
University College London - Gower Street - London - WC1E 6BT Tel:+44 (0)20 7679 2000

© UCL 1999–2011

Search by