Behavioral Interventions from Trait Insights

Abstract

Individuals have the stated and unstated beliefs and intentions. The theory of planned behavior is expressed by the mathematical function where beliefs have empirically derived coefficients. However, personality traits can help account for differences in beliefs. In this paper, we will find out how we can amplify behavioral interventions from text-based trait insights. Therefore, we research techniques (e.g., sentence and word embedding) behind text-based traits. Furthermore, we exemplify text-based traits by 52 personality characteristics (35 dimensions and facets of Big Five, 12 needs and five values) and 42 consumption preferences via API of the IBM Watson™ Personality Insights service. Finally, we discuss the possibilities of behavioral interventions based on the personality characteristics and consumption preferences (i.e., text-based differences and similarities between the individuals).