University of Dublin, Trinity College
Post Title: Research Fellow
Status: 3 year contract
Dept/Faculty: School of Computer Science & Statistics
Closing Date: Until position is filled
Salary: 32,358 - 38,526 euro per annum
(Position 1)
You will work on the extraction of `sentiment' bearing sentences
from continuous news streams. The impact of news on shares, currencies
and other financial instruments is discussed under the umbrella of
news impact analysis in econometrics and sentiment analysis in
behavioral finance. `Sentiment' extraction requires the
development of algorithms that are informed by methods of information
extraction, corpus linguistics and by local grammar formalism. This
information is needed particularly by the financial services industry,
political and cultural analysts, and by public at large.
Your role will be to develop and implement algorithms that will
facilitate sentiment analysis. You will test the usefulness of the
algorithms by building computing systems for analyzing real world news
streams as provided by news agencies across the world. You will
interface your system to a time-series analysis system that analyses
the behaviour of financial markets purely on the basis of numerical
information. Your work will clarify research questions in the emerging
multi-disciplinary field of sentiment analysis. The clarified research
questions will help in the systematic development of information
extraction systems which will be at the heart of the so-called
semantic web.
(Position 2)
You will work on the analysis and synthesis of information in two
modalities: images and texts, and numbers and texts. Such information
is communicated in everyday interaction and humans are very adept at
annotating images (and streams of numbers) with textual descriptions
and equally at illustrating verbal descriptions. This work requires
development of algorithms that are informed by information extraction
and image processing. You will build a neural computing system that
learns to deal with cross-modal information and will be able to
auto-annotate and auto-illustrate. The annotation and illustration are
critical for the work of those running large image libraries --
health care, forensic services, and indeed the Web is filling up with
images very fast.Your role will be to develop and implement the
algorithms, and test the usefulness of the algorithms by building
computing systems for analyzing real world video streams. You will
interface your system to an attention modeling system that simulates
how humans focus on `objects' of interest in a visual field and
then describe the objects in a succinct manner. Your work will clarify
research questions in this emerging multi-disciplinary field of
cross-modal analysis. The clarified research questions will help in
the systematic development of information extraction systems which
will be at the heart of the so-called semantic web. The questions will
be of equal interest in neurosciences where it is now possible to
observe human brain as it processes information in various modalities
through functional magnetic resonance imaging.
Qualifications for both:
You will have a PhD or equivalent in computing, computational
neuroscience or cognitive science. You will have a strong background
in information extraction, ontology and terminology. An interest in,
and familiarity with, computer vision methods and techniques is
desirable. The appointment will be for a fixed period of three years.
Candidates should submit a full curriculum vitae, to include the names
of three referees, to:
Khurshid Ahmad
Professor of Computer Science
Trinity College
Dublin 2
Tel: +353 1 608 8429
Fax: +353 1 677 2694
Email:khurshid.ahmad@cs.tcd.ie
We welcome applications by e-mail. Please note there is no application
form to be completed.
TRINITY COLLEGE IS AN EQUAL OPPORTUNITIES EMPLOYER
This archive was generated by hypermail 2b29 : Tue Apr 11 2006 - 12:44:37 MET DST