| More Information: | The Department of Informatics at the University of Sussex invites applications
for a PhD studentship working in the Text Analytics Group under the supervision
of Julie Weeds, in the area of compositional distributional semantics. Expected
start date September 2017. Application deadline 27/1/2017.
Traditional distributional similarity methods and more recent work on
distributed representations of words learnt by neural networks share the view
that a word can be represented by a vector which is determined from the contexts
in which it is used. Using the distributional hypothesis, it is then often
assumed that words which are distributionally similar are also semantically
similar i.e., similar in meaning. Applications which make use of knowledge about
the similarity of words include speech recognition, parse selection, information
retrieval, information extraction, question-answering and natural language
generation. However, important challenges remain in this area.
This project will aim to address at least one of these, which is how to move
from similarity at the word level to similarity at the phrase and sentence
level. The student would be expected to contribute to the development and
evaluation of syntax-sensitive compositional distributional semantic models for
use in the application of Question Answering.
The studentship will include a three year stipend at a standard rate (currently
£14,296 per annum) and, in addition, fees at the UK/EU rate. Since the
studentship only covers fees at the UK/EU rate, overseas applicants are kindly
requested to state in their application how they propose to cover the difference
between the UK/EU rate and the overseas fees (for more details visit:
http://www.sussex.ac.uk/study/phd/fees-and-scholarships).
For more details about how to apply see:
http://www.sussex.ac.uk/study/phd/fees-and-scholarships/scholarships/view/704
For informal enquiries about the project - contact: Julie Weeds
(j.e.weeds@sussex.ac.uk) |