James Barry is a Research Scientist at IBM Research Europe, and is based in Daresbury, UK. His work focuses on utilising Natural Language Processing (NLP) techniques to aid with materials discovery. In particular, James is focusing on a number of methods for extracting structured and unstructured information from large collections of scientific texts, which can be used to train foundation models and also be included in question answering systems and knowledge graphs.
James is interested in many NLP applications, including training foundation models, information retrieval, knowledge graph creation, and question answering. Prior to joining IBM Research, James completed a PhD on the topic of multilingual dependency parsing at the ADAPT Centre, Dublin City University.