Empowering scientific discovery through AI and data-intensive computing
3 July 2024
The research of EPCC Chancellor’s Fellow Rosa Filgueira focuses on pioneering solutions that streamline the extraction of actionable knowledge from vast datasets, thereby accelerating scientific discoveries. As she explains here, she is thrilled to embark on an innovative journey that is using AI, data science, and data-intensive computing to redefine scientific possibilities across diverse fields.
Bridging the computational gap
In today’s rapidly evolving technological landscape, numerous scientific disciplines are becoming increasingly reliant on data-intensive methods, and the automated analysis of expansive datasets is now a cornerstone of the scientific process. However, the computational demands associated with this analysis can be daunting, particularly for non-expert users. My research aims to bridge this gap by developing intelligent adaptive systems that empower users to create, discover, and share data-intensive computing applications effortlessly.
Key research areas
My research will focus on several key areas over the course of my Fellowship:
- Adaptive optimisation techniques: Developing systems for domain experts to use data-intensive methods without needing deep platform knowledge.
- Semantic repository insights: Enhancing software repository analysis with deep learning techniques for cross-repository code comparison and exploration, improving efficiency in scientific software development.
- Innovative code exploration: Utilising natural language processing (NLP) to develop tools for semantic code search, summarisation, and completion, simplifying code reuse and improving software development processes.
- Text mining and NLP within digital humanities: Advancing text mining, NLP, and deep learning techniques to extract insights from historical texts, analyse cultural events data, and enhance democratic governance.
A holistic and distinctive approach
My approach is distinctive in that it combines multiple aspects, such as new programming abstractions, auto-parallelisation techniques, resource elasticity, and machine learning for source code, to develop a new generation of ‘smart’ data- and compute-intensive applications. This holistic approach will result in end-to-end solutions tailored to various scientific disciplines, feeding into the development of general domain-independent frameworks and techniques.
Collaborative potential
The benefits of my research extend across numerous scientific and business domains, including seismology, economics, climate science, social computing, health, urban studies, and the digital humanities. By fostering collaborations within the University of Edinburgh and beyond, I aim to unlock the capabilities of advanced computational techniques for a broader range of non-expert users.
Conclusion
The Chancellor’s Fellow opportunity at EPCC aligns perfectly with my research aspirations and collaborative mindset. By driving innovation in AI and data-intensive scientific computing, I aim to enhance productivity, foster the adoption of reproducible and reusable scientific software, and ultimately accelerate data-driven scientific discovery and business intelligence.
Chancellor's Fellowship scheme
The Chancellor’s Fellowships scheme at the University of Edinburgh is a prestigious five-year programme aimed at guiding outstanding early career researchers towards becoming leaders in their scientific fields and driving innovation. You can read about the work of our other Fellows on our website:
- Oliver Brown
Article: Chancellor’s Fellowship in Quantum Computing - Joe O'Connor
Article: Quantifying uncertainties in direct numerical simulations of a turbulent channel flow