Automating knowledge synthesis in biomedical literature using AI and Large Language Models

Project Code

MRCPHS25Br Liu

Research Theme

Population Health Sciences

Project Summary Download

Summary

There are millions of new scientific findings published every year, and the evidence they contain vary by their study designs and analysis methods (e.g. randomised trials, Mendelian randomisation studies, observational studies, etc). The student will use artificial intelligence (AI) methods such as Large Language Models (LLMs, e.g. LLaMA and GPT-4) for knowledge synthesis, i.e. to extract and summarise from large amounts of heterogenous information to generate broader perspectives and new insights, to be able to answer questions such as the effect of obesity on breast cancer as identified by the vast volumes of literature.

Lead Supervisor

Dr Yi Liu

Lead Supervisor Email

yi6240.liu@bristol.ac.uk

University Affiliation

Bristol