Leveraging language models to categorize individual‐level priorities and deliver personalized brain health recommendations

There is heterogeneity in what matters to individuals in the context of cognitive health. Person-specific priorities should be taken into account when planning treatments. We show that using language models, we are able to identify clusters which most closely align with individual free form responses. By creating educational materials and behavioral interventions tied to these clusters, we can eventually provide personalized recommendations at scale to sustain brain health and reduce the risk of brain-related disabilities (e.g. dementia).