Interactive Seminar

Day 1: Tuesday, 13 September


Natural language processing to capture person-first or identity-first language

Session convener: Julia Kasmire, University of Manchester

The presenters are looking to develop a collaboration using text-mining and other natural language processing methods. They want to capture the context around how practitioners and health researchers use person-first or identity-first language in their documents. For example, do they tend to use "people with autism" or "autistic people"? Do they speak the same way about children and adults? Has the language used changed over time? Are there links to official publications of guidance or are the patterns driven from the bottom up? Researchers may have some intuitions about how language is used and what this may mean for practitioners or health care users. But intuitions are only the first step to using advanced methodological techniques to evidence the patterns and discern the factors at play. We will introduce the concepts, present some preliminary results, foment discussion, and solicit feedback or potential collaborators. We want to allow audience members to share their own intuitions and to reflect on how they came by those intuitions. We also want to learn how those intuitions might be investigated as research hypotheses, and how such research can challenge ideas that would otherwise have been accepted uncritically.