Emotion Analysis from Text

Emotions are a concept that we all know and that feels familiar. Nevertheless it’s quite hard to define it, and making computer to understand what an emotion is requires to formalize the concept. In this class, we will learn about emotion theories from psychology and work in natural language processing/computer science that formalizes them into computational models. We will learn how to build automatic emotion detection models and how to apply them, for instance for social media analysis, computational literary studies, or computational social sciences. We focus on text as the only modality in the class.

There is no dedicated book on emotion analysis, but some material can be helpful:

• The book on affective computing by Rosalind W. Picard. This book is broader than this class and discusses emotions and affect as it is modelled with computers in more general.
• The book on sentiment analysis by Bing Liu. This book focuses on text analysis, but more on opinion mining and sentiment analysis, less on emotions.
• The handbook of emotions by Feldman Barrett, Lewis, Haviland-Jones. This book provides a very good overview on the theories of emotions from a psychological perspective.