Online emotions and how to measure them

Abstract

Our online life generates a vast amount of digital traces, for example when we use our phone or talk with others on social media. These traces can be used to study aspects of emotional life that were very difficult to study with experiments or surveys. I will present work on the development and application of digital trace data analysis to study emotions at various levels. I will present LEIA, a language model to identify individual emotions from social media text and how text analysis methods can be applied to create emotion macroscopes of society at large.