Information Retrieval
The Glasgow Information Retrieval Group was founded in 1986 by Professor C. J. ‘Keith’ van Rijsbergen, often considered one of the founders of modern Information Retrieval (IR). From its outset, the Glasgow IR group has focused on improving the effectiveness of IR systems, inventing new logic & probabilistic retrieval models in the 90's and early 2000's, followed by the development of adaptive query expansion techniques, interactive multimedia models, the Divergence From Randomness framework, and leading research in quantum IR, expertise search, and search result diversification. More recently, the group has embraced machine learning, deep learning and generative AI technologies for search, recommendation and information access at scale, remaining at the forefront of research that combines effectiveness, efficiency, and real-world impact.
As one of the longest-running and most research-active Information Retrieval groups in the world, and the most prolific in Europe by publication output, we continue to advance the state of the art across the full spectrum of IR research and are always keen to hear from from outstanding prospective PhD students. We actively welcome collaborations with academic and industrial partners, and are always keen to engage with organisations seeking innovative solutions to information access, search, recommendation, and AI-driven discovery challenges.
Our research spans the full range of topics relevant to the development of modern IR systems, including: Large-scale IR & Efficient IR, Recommender Systems Models, Responsible AI & IR, Natural Language Processing, Data Streams, Applications and Evaluation.
Project Highlights
PyTerrier
PyTerrier is a leading open-source Python framework for information retrieval (IR) research, experimentation and teaching, developed at the School of Computing Science. Building on decades of experience from the established Terrier IR platform, PyTerrier provides an elegant, declarative way to rapidly build, combine and evaluate state-of-the-art retrieval pipelines. Its extensive ecosystem supports hundreds of IR datasets, a wide range of evaluation measures, and techniques spanning classical retrieval, learning-to-rank, dense and learned-sparse retrieval, neural re-ranking and RAG. With a large international user community, PyTerrier enables rapid development of new retrieval approaches and reproducible research and won Best Search Project at the BCS Search Industry Awards 2025.
CLIMINVEST
ClimInvest is a collaborative European research initiative bringing together climate scientists, sustainability experts and financial stakeholders, including researchers from the University of Glasgow, to bridge the gap between climate science and investment decision-making. The project develops tailored tools, indicators and maps to help investors assess physical climate risks, such as flooding, sea-level rise and heatwaves, while supporting the development of financially viable climate adaptation projects. Through its focus on bankable-by-design (BbD) solutions, ClimInvest translates complex scientific data into practical, sector-specific insights that enable asset managers, banks and public authorities to evaluate, fund and implement effective adaptation strategies. By combining technical innovation, alignment with ESG frameworks and stakeholder collaboration, the initiative aims to reduce barriers to investment and accelerate the deployment of high-impact, climate-resilient projects across Europe.
PHAWM
The Participatory Harm Auditing Workbenches and Methodologies (PHAWM) project is a University-led initiative designed to make Artificial Intelligence safer and more trustworthy. Backed by £3.5 million from Responsible AI UK (RAi UK), it focuses on "democratising" AI auditing by creating open-source tools that non-experts can use to identify biases or harms in AI systems.