This article explores the automatic recognition of personality traits in conversation and text, focusing on the Big Five personality dimensions: extraversion, emotional stability, agreeableness, conscientiousness, and openness to experience. The authors use both self-ratings and observer ratings of personality to train statistical models, including classification, regression, and ranking models. They analyze the impact of different feature sets on model accuracy and find that ranking models generally perform best. Models trained on observed personality scores outperform those trained on self-reports, and the optimal feature set varies depending on the personality trait being modeled. The study also reveals new linguistic markers for personality traits, confirming previous findings and providing insights into the relationship between language and personality.This article explores the automatic recognition of personality traits in conversation and text, focusing on the Big Five personality dimensions: extraversion, emotional stability, agreeableness, conscientiousness, and openness to experience. The authors use both self-ratings and observer ratings of personality to train statistical models, including classification, regression, and ranking models. They analyze the impact of different feature sets on model accuracy and find that ranking models generally perform best. Models trained on observed personality scores outperform those trained on self-reports, and the optimal feature set varies depending on the personality trait being modeled. The study also reveals new linguistic markers for personality traits, confirming previous findings and providing insights into the relationship between language and personality.