The GRILL team consists of School of Computing Science graduate students Carlos Gemmell, Federico Rossetto, Iain Mackie, Sophie Fischer, and Paul Owoicho with Dr. Jeff Dalton as faculty advisor.
The GRILL team consists of School of Computing Science graduate students Carlos Gemmell, Federico Rossetto, Iain Mackie, Sophie Fischer, and Paul Owoicho with Dr. Jeff Dalton as faculty advisor. GRILL will be able to help people solve complex tasks in the kitchen, garage, and beyond. We will develop novel multi-modal algorithms that are grounded in the real world that can understand what you're doing, how you're doing it, and adapt to new challenges on the fly. This research will focus on reasoning over past interactions, modular structured task representations, and multi-modal information extraction.
Carlos G. - Team leader
Carlos is a second year NLP PhD student at the University of Glasgow supervised by Jeff Dalton. He is interested in developing neural language systems capable of externalising knowledge and computation through use of structured tools for question answering and conversational information seeking, as well as measuring the generalisation, robustness and reasoning capabilities of such systems. He has been published in SIGIR and led the GRILL Lab at TREC 2020 in the top five teams for conversational response ranking. Before his PhD, Carlos earned his B.S. degree in Computer Science from the University of Glasgow and interned as a Vision and Machine Learning engineer at HiBirdi.
Paul is a first-year PhD student studying the role of mixed-initiative in effective conversational search. He is particularly interested in how search systems can meet a user's complex information needs by expertly guiding them through the information space. Paul earned his MSc at the University of Glasgow where he developed novel Transformer-based summarization methods for the TREC Podcasts track. Paul also developed retrieval and query rewriting baselines for the TREC 2021 CAsT Track that involved cloud deployment on AWS with Docker, Kubernetes, and multi-GPU micro-services. In his free time, Paul enjoys weightlifting and cooking Nigerian recipes.
Sophie is a 4th-year BSc student, starting a research Master’s (MSci) in the Fall. Her Honours thesis project “COP Bot: An agent for climate change education and advocacy” is a collaboration with the BBC Voice + AI group and received the best thesis award. She developed a multi-modal (voice + screen) Dialog-flow based agent deployed on Google cloud. It provides rich functionality with question answering, event recommendation, and social chit-chat elements based upon a specially crawled corpus around COP26 and climate change. She will assist in UX development, perform user lab studies, and develop novel approaches to enhance user engagement.
Federico is a second-year PhD student at the University of Glasgow working on Multimodal Learning for conversational systems, integrating audio, video and text. He graduated from the University of Pisa in Computer Science following an AI curriculum in 2019. Federico won the Fujitsu AI NLP Challenge 2018 with his team.
Iain previously worked as a Quantative Trader, founded an e-commerce start-up that supports one of the UK's fastest-growing brands, and is an Investor at Creator Fund (an early-stage European VC). He holds an MA in Applied Economics from the University of St Andrews and an MSc in Data Science from the University of Edinburgh. Iain is currently pursuing a PhD at the University of Glasgow under Jeff Dalton. His PhD focus is multi-task deep learning for information extraction and knowledge-centric ranking tasks. Iain develops new methods of retrieval for complex tasks using knowledge bases and text-based graph neural networks.
Jeff Dalton - Faculty advisor