Claire’s Journey: Exploring ways to inform, speed up and improve the legal workflow using NLP
“Legal language provides an interesting framework for exploring key challenges in NLP, such as extracting salient information from text and understanding how current language models process non-generic text with distinct structural and linguistic properties.”
Based between the School of Informatics and the University of Edinburgh Law School, Claire Barale has spent her PhD exploring the intersection of legal reasoning and Natural Language Processing (NLP). Interested in knowledge-intensive tasks specific to the legal domain, her research focuses on designing NLP-based tools to enhance the transparency, efficiency and quality of legal workflows.
Claire was drawn to this area of research because she recognised the important role that language plays in enhancing the quality of current AI systems.
“Language is essential for many applications, such as helping people navigate legal processes and providing instructions to large language models (LLMs). Reliable natural language understanding is important for all of these tasks, and its enhancement is one of the keys to advancing AI systems.”
Claire’s work also highlights the unique challenges legal language presents for NLP models: “Legal language is a distinct linguistic register with unique challenges for NLP. Simply increasing pretraining or post-training data is not always effective, especially given the scarcity of high-quality legal text and proprietary restrictions.”
As part of her PhD project, Claire has specifically been examining this in relation to the refugee claim process in the Canadian context. She has explored different approaches to the design of an ethical human-AI reasoning support system for decision-makers in refugee law, as well as examining the possibility of using machine learning to help reduce the influence of external factors (which unavoidably impact human decision makers) in refugee law decision making.
Reflecting on her PhD journey, Claire has enjoyed being part of the Centre’s interdisciplinary community throughout her studies. She has found the opportunity to receive feedback on her research project particularly beneficial due to the diverse disciplinary backgrounds of students and staff.
Among the many highlights of her PhD so far, Claire is particularly proud to have received the Best Doctoral Consortium Paper Award at the International Conference on Artificial Intelligence and Law (ICAIL) 2023. She was also one of just nine early-career researchers awarded the Bloomberg Data Science PhD Fellowship for 2023-2024! As part of this prestigious fellowship, Claire spent four months at Bloomberg’s New York office for a Research Scientist Internship, where she worked with the AI and Law team, focusing on reasoning about temporal structures in legal texts.
Now nearing the completion of her PhD, Claire is looking forward to the next chapter in her career. She is hoping to continue advancing research in information extraction and commonsense reasoning with language models. Whether through a research scientist role in industry or a postdoctoral role in academia, Claire is actively seeking opportunities that will enable her to strengthen her expertise and contribute to solving broader challenges in natural language understanding.
We are proud of Claire’s achievements and the contribution her work is making at the intersection of law, AI and ethics.