[Corpora-List] PhD studentship available: Interaction Mining
Matthew Purver
m.purver at qmul.ac.uk
Fri Mar 25 15:01:02 UTC 2011
One funded PhD studentship in Interaction Mining is available with the
Interaction, Media and Communication research group at Queen Mary
University of London. Suitable candidates with experience in natural
language processing, information retrieval and/or dialogue processing
are encouraged to apply - please contact Matthew Purver
<m.purver at qmul.ac.uk> or Tassos Tombros <tassos.tombros at eecs.qmul.ac.uk>
with enquiries.
Search engines let us find information in large collections of data -
but only when we can express what we're looking for in terms of its
content (for web pages, the words of interest; with the semantic web,
perhaps their meanings). What if we're looking for patterns of
interaction instead? We often want to know about particular kinds of
events in conversation: decisions made in business meetings, answers to
questions in radio broadcasts, agreements and disagreements in social
media. In this case, keywords or semantic categories can't help us: we
need to look for the characteristic structures in the interaction itself
(e.g. the characteristic patterns of the decision-making process). We
often need to do this with the errorful output of speech recognition,
the unconventional language and spelling used on social media, or the
rich combination of verbal and non-verbal information in face-to-face
conversation. This project will investigate the combination of
rule-based models of dialogue structure with robust statistical methods
to extract useful information and summaries from naturally occurring
interaction data.
More information about PhD studies at QMUL and the Interaction Media and
Communication research group is available here:
http://www.eecs.qmul.ac.uk/phd/
http://www.eecs.qmul.ac.uk/phd/group.php?group=imc
http://www.eecs.qmul.ac.uk/phd/apply.php
--
Matthew Purver - http://www.eecs.qmul.ac.uk/~mpurver/
Lecturer in Human Interaction
Interaction, Media and Communication Group
School of Electronic Engineering and Computer Science
Queen Mary University of London, London E1 4NS, UK
_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora
More information about the Corpora
mailing list