ASAP - Automatic Social Assessment and Prediction
Judging others is a ubiquitous and consequential phenomenon. Social judgments color our professional and private life and guide relevant social decisions. Understanding how social judgments are formed has long been a key topic of psychological research. What physical and behavioral cues are most relevant and how are they integrated by perceivers to form social judgments? To date research on social judgment-making has focused on selective sets of extracted cues and restricted statistical models of cue integration. In this project we propose to complement the existing theory-driven research by applying data-driven approaches in cue extraction and integration using state of the art machine learning models to overcome limitations of manual cue extraction and traditional integration models.
We investigate judgments in two highly relevant selection contexts: Personnel and Mate Selection. Making judgments on whom to select as a future employee and a future romantic partner, respectively, represent two of the most consequential social tasks in occupational and private contexts or in words purported to be said by Freud: “Love and work, work and love – that’s all there is.” In both contexts, the power to predict resulting judgments is currently limited. Also, in both contexts, there is currently a lack of understanding with regard to how exactly (i.e., by means of what sort of mental integration of what sort of observable cues) judgments are made. We aim to overcome these limitations with our approach of automatic cue extraction and computer-based integration with the goal to thus create a better understanding of social judgments in these domains.