PhD scholarship on Data Science and Machine Learning

University College London, UK
Sunday, May 1, 2016

One full PhD scholarship on Data Science and Machine Learning from Computer Science, University College London

UCL Computer Science has a full PhD scholarship available on topic “Conversational Search: Mathematical Modelling and Applications”. It is funded by the prestigious Microsoft Research PhD scholarship programme.

The PhD student will be jointly supervised by Dr. Jun Wang at Computer Science, University College London (UCL) and Dr. Filip Radlinski  at Microsoft Research Cambridge, UK. The student will be based at UCL, but periods of study at Microsoft Research Cambridge are expected.

The candidate should have a BSc and preferably an M.Sc. in Computer Science or related subject, and have previous work or research experience on Reinforcement Learning, Deep Learning, or Natural Language Processing. We expect the candidate to have strong math skills and is familiar with statistical modelling.

Topic (subject to the discussion): Conversational Search: Mathematical Modelling and Applications

Summary: Many complex information retrieval tasks involve conversations and explorations. These range from simple cases where the same user fails to obtain suitable results in an earlier interaction and keeps refining queries, to complex exploratory search scenarios where no single item or single query suffices and users explore the information space sequentially. We see in many cases like personal assistant (e.g, Cortana, Siri), dialogues might be needed in
order to provide personalised content and services. This PhD project aims to develop theoretical groundwork for such a multi-interaction retrieval and recommendation setting, and then measure the effectiveness of resulting algorithms in a practical multi-interaction system. This will deliver a principled and effective Conversational Search (CS) framework. This is data driven research that bridges machine learning, natural language processing, and information retrieval.

Interested candidates should send an email to Dr. Jun Wang (