Postdoc in statistical NLP/ML for source code text

Postdoc in statistical NLP/ML for source code text

Posted by Rebecca Martin on Mon, 17/06/2013 - 13:47

The School of Informatics at the University of Edinburgh is seeking an excellent postdoctoral researcher for a new EPSRC-funded project to apply methods from statistical natural language processing and data mining to find patterns in computer program source code. The project is supervised by Dr Charles Sutton.

Modern software developers often find it necessary to use software libraries and programming languages with which they are unfamiliar, which can lead to longer development times and lower reliability. However, on the Internet billions of lines of open source code are readily available, code that contains a large amount of implicit knowledge about good coding practice. The goal of the project is to find patterns in large corpora of source code text, which can be used to help developers to write better code, in effect transferring knowledge from experienced developers to less experienced developers.

The researcher will be a part of the School of Informatics at the University of Edinburgh. This is an opportunity to work in world-leading groups for machine learning and NLP. More broadly, a recent international review described the School as an "elite" department of computer science in Europe, and in national research assessment exercises, the School of Informatics has consistently ranked at the top in the UK for research quality.

The successful candidate will have a background in statistical language processing or machine learning and a strong interest in the application area. We will also consider applications from researchers in software engineering who have strong interest in building their knowledge of techniques from statistical NLP and machine learning.

For informal enquiries, please contact Charles Sutton csutton(at)
For full consideration, please apply by 19 JULY 2013.

For more information, see