Dimensions of Semantic Change: Applying the SIBling Framework to Mental Health Concepts

It was a pleasure to present my PhD work at the Natural Language and Text Processing Lab, Utrecht University. I introduced two complementary frameworks for studying lexical semantic change: (1) SIBling — models change along Sentiment, Intensity, and Breadth; and (2) LSC-Eval — generates synthetic benchmarks for evaluating methods to detect change. Together, they offer new tools for tracing socially significant conceptual shifts, with applications to mental health concepts. Talk info: https://nlp.sites.uu.nl/2025/09/11/nltp-content-meetings-september-16-dimensions-of-semantic-change-applying-the-sibling-framework-to-mental-health-concepts-by-naomi-baes/ | Slides Above.