Phd Position - Adaptive Natural Language Generation
I am happy to announce that I have funding for a PhD position at the university of Aberdeen CS department. I am keen to recruit a motivated and creatively-minded student to work with as we explore, evaluate and extend the potential of LLMs for generating more audience-aware, adaptive, and ultimately more human-like language. This is an exciting position, as there is a degree of freedom for us to decide the direction of the project together (You can get an idea of the directions I'm interested in from my publications). The official advert can be found on find a phd, however I provide some further details below.
Project Outline
Are you interested in the intersection of natural language processing, psycholinguistics and cognitive science, specifically as it applies to large language models (LLMs) and natural language generation (NLG)?
This project is recruiting a PhD student to research how LLMs can produce more human-like language that is adaptive, context-specific, and audience-aware.
Current dialogic interaction systems fall short when it comes to nuanced, pragmatic language understanding. Taking inspiration from the fields of psycholinguistics and cognitive science, this project aims to explore, evaluate and extend the potential of LLMs to generate language that is appropriate for an interactive setting such as dialogue.
Adaptive NLG can lead to more inclusive, personalised, and appealing language; essential components of a number of downstream applications, such as in educational settings, or as creative tools. Evaluating such adaptive language will involve analysis of language produced by humans, compared to humans and models in collaborative communicative settings.
The output of this PhD project should advance the state of the art in NLG systems, in terms of techniques for evaluation, understanding of LLMs, and/or improved generation techniques - thus allowing systems to adaptively interact in a manner that is sensitive to linguistic contextual cues. Applicants will be expected to work on a topic within the area outlined above, aligned with the research interests of Dr Arabella Sinclair.
Supervision
Typically, you will meet with your primary supervisor on a weekly basis, at least until you reach the stage of writing up your thesis, or during periods where an internship is undertaken.
Ehud Reiter will act as the second supervisor for this PhD position. His expertise in Natural Language Generation and Evaluation will be a valuable addition when it comes to receiving feedback on some of your work.
If you have further questions, please contact me. Ideally please add [phd-adaptNLG] to the subject line of initial email correspondence, so that I can manage applications more easily.
Milestones
Your work will be assessed at the end of year 1, where you will be evaluated on your progress by two assessors (other than your supervisor) both via oral interview and on a progress report that you prepare. If after this assessment, your work is found to be on track for successful PhD completion, you will continue with the program, and will be expected to produce a substantial original thesis within at most 4 years from the date of admission. To successfully obtain a PhD in Computing Science from the University of Aberdeen, you will need to defend your thesis orally to two appointed examiners.
Candidate Requirements
Selection will be on the basis of academic merit. The successful candidate should have, or expect to obtain, a UK Honours degree at 2.1 or above (or equivalent) artificial intelligence, computer science, machine learning, computational cognitive science, or a related discipline along with strong programming skills and demonstrable previous experience in natural language processing. Lastly, the candidate should have a keen interest in the area and enjoy working collaboratively.
Applications from candidates belonging to demographics typically underrepresented in AI are particularly encouraged to apply.
Funding
This is a studentship open to both home and overseas students funded by the School of Natural and Computing Sciences. Funding will cover tuition fees for 36 months (3 years), paid monthly in arrears. The start date is to be agreed with the supervisor and we reserve the right to remove the advert if a suitable candidate is found before the advertised closing date
Application Process
The official application process is outlined on the advert on find a phd. Please get in contact with me if you have any questions about this. Your cover letter should act as a demonstration that a) you are able to write well and concicely, b) you are interested in this project specifically rather than some other topic c) that you consider yourself a good fit with the project and supervision team. The Project proposal is also a demonstration that you are a) keen enough on this area to read some related papers (I suggest starting from my publications and projects to see whether they are interesting then starting from there) and b) able to highlight some initial starting points to your research in the first year or so of the project that you are most interested in (this is very much allowed to change).