PhD Affiliate Research Spotlight

Each term members of the Data Science Institute’s PhD Affiliate Network are invited to share their research at our Research Spotlight events. Our speakers this month are Hannah Bunt and Zarja Hude.

Bridging fragmented literatures: Using large language models (LLM) to navigate conceptual chaos in systematic reviews

Hannah Bunt, Department for Psychological and Behavioural Science

Research across disciplines struggles with a fundamental problem: the same words often describe different things, whilst different words describe the same phenomena. This talk demonstrates how LLMs can help systematic reviews overcome this 'jingle-jangle' problem by identifying relevant research across fragmented literatures that don't realise they're studying the same thing. Drawing on a systematic review of NLP-based studies of complaints that screened over 300,000 papers, Hannah shows how AI can navigate terminological ambiguity at massive scale, offering a scalable approach for any field where researchers unknowingly work in isolated silos because they use different words for the same ideas.

Neuro-Symbolic Legal Reasoning

Zarja Hude, LSE Law School

LLMs have shown considerable promise in legal tasks yet trust in their use remains limited due to non-determinism and opacity – qualities that conflict with the transparent, interpretable, and auditable nature of legal reasoning. This PhD project explores how neuro-symbolic methods, which combine neural networks with logic-based systems, can help bridge this divide, with a designed backed in legal philosophy. Implementation results demonstrate that LLMs perform significantly better in legal reasoning when guided and constrained by symbolic components – a principle that may extend to any domain requiring interpretable, high-stakes reasoning.

About the DSI PhD Affiliate Network

We invite doctoral researchers, from all departments across LSE, who have an interest in data science and artificial intelligence to join the DSI PhD Affiliate Network; whether you are interested in applications for your research, are studying the social scientific implications of AI and data or are working on fundamental methodologies, please sign up.

LSE holds a wide range of events, covering many of the most controversial issues of the day, and speakers at our events may express views that cause offence. The views expressed by speakers at LSE events do not reflect the position or views of the London School of Economics and Political Science.