From Valence to Emotions: How Coarse versus Fine-Grained Online Sentiment can predict Real-World Outcomes
- 80 páginas
- 3 horas de lectura
Focusing on the intersection of user-generated content and scientific research, this book explores how emotions and valences in online data can predict human-related events. It begins with a theoretical framework for sentiment detection and classification, then categorizes empirical literature into three prediction subjects. A thorough analysis compares various prediction methods, assessing their consistency and limitations, providing insights into the effectiveness of these approaches in understanding human behavior through digital expressions.
