Analysis of personal attribute relationships based on online research and social data

抄録

Although social data can be an important source of information for studying personal attributes, it remains unclear which kinds of attributes can be inferred from social data and meaningfully used in the social sciences. In response, we analyzed the relationships between a variety of attributes, including political alignment, morality, food identity, and brand preferences, based on social data (i.e., Twitter text) and online survey data. Our results revealed a relationship between political alignment (liberal and conservative) and morality, food identity, and brand preferences. In addition, liberal people were more likely to belong to the food left-wing category. As for brand preferences, the conservative people showed a higher preference for Japanese companies than liberal people.