Understanding the Level of Trust in Twitter based on the three Psychological Facets - Predictability, Dependability and Faith

Trust plays a significant role in social relations. Automatic trust prediction has become an important research area recently. However, trust-prediction is still an ill-defined problem. There is no concrete definition of trust that can be found in literature, mostly it is task-specific so far. In this paper, a generic definition of trust has been borrowed from psychology, which includes three facets of trust namely predictability, dependability, and faith. To understand the role and influence of these facets we have crowd-sourced a Twitter network data through Amazon Mechanical Turk and Rapid Workers. People were asked to rate their followers/followings (aka friends) on a Likert scale ranging from 0 to 5 (0 as low and 5 as high) into these three facets. Then to understand ` who trusts whom' in a social network user level personality (aka person level sentiment) and their societal traits in terms of values (societal sentiment) has been analyzed. For the personality analysis, the Big 5 or the OCEAN model (Openness, Conscientiousness, Extroversion, Agreeableness, and Neuroticism) has been used and to understand the sociological sentiment the Schwartz values (Achievement, Benevolence, Conformity, Hedonism, Power, Security, Self-Direction, Stimulation, Traditional, and Universalism) has been utilized. Empirical results reveal several interesting insights of trust relations in societal relations.