A modelling framework for collective emotions in online communities

David Garcia ,  Antonios Garas ,  Frank Schweitzer 

ETH Zurich, Weinbergstrasse 56/58, Zurich 8092, Switzerland

Abstract

Collective emotions in online communities can be triggered by external events, but also can emerge from the interaction of many individuals. These collective emotional states can appear faster and have different properties than in offline interaction.  The emotional expression left by millions of Internet users allows us to quantitatively study emotions at a large scale, but the dynamical explanation for the way they emerge and evolve requires additional work.  We present our modeling framework for collective emotions in online communities.  This framework allows us to design and analyze agent-based-models that reproduce collective emotions in different online communities.  These models provide a link between macroscopic emotional behavior, and the dynamics of psychological states under microscopic user-to-user interaction.  We empirically test the assumptions of this framework in experimental setups, measuring individual emotion dynamics through physiological signals, and self-assessment of emotions.  We illustrate two applications of our framework to two very different online communities.  Our first model focuses on emotions in product reviews communities, reproducing the empirical distribution of emotions towards products in Amazon.com. The second model reproduces the emergence of collective emotional persistence and long-range interactions in real-time IRC chatrooms. We summarize by discussing further applications of our approach to model emotions in Open Source Software projects.

 

Presentation: Oral at CyberEmotions conference, by David Garcia
See On-line Journal of CyberEmotions conference

Submitted: 2012-12-03 13:23
Revised:   2012-12-03 13:28