Keynote 1: NLP for Personality Assessment in Life Narrative Interviews by Dr. Josh Oltmanns

Knowledge of personality structure and resulting assessments stem from the study of language. Advances in artificial intelligence (AI) offer a promising avenue towards improving assessment and understanding of personality. A representative sample of N = 1,405 older adults in St. Louis (33% Black, 65% white) completed a life narrative interview, the SIDP-IV, NEO-PI-R, and self-report measures of physical functioning and depressive symptoms. Language associated with personality and personality disorder was modeled in three ways: 1) Parameters from the RoBERTa language model were fine-tuned, 2) Topics in the life narratives were modeled with BERTopic, and 3) other features of language were extracted using LIWC. Features from were then combined in feed-forward neural networks to create language models in each domain, which were then combined in a linear regression to create multimodal models of personality traits. Multimodal language models were the validated through cross-sectional correlations with self-reported functioning measures. Findings demonstrate promise of language-based AI with exciting future opportunities to provide automatic personality assessment and prediction.