Abstract

Communication within online social network applications enables users to express and share sentiments electronically. Existing studies examined the existence or distribution of sentiments in online communication at a general level or in small-observed groups. Our paper extends this research by analyzing sentiment exchange within social networks from an ego-network perspective. We draw from research on social influence and social attachment to develop theories of node polarization, balance effects and sentiment mirroring within communication dyads. Our empirical analysis covers a multitude of social networks in which the sentiment valence of all messages was determined. Subsequently we studied ego-networks of focal actors (ego) and their immediate contacts. Results support our theories and indicate that actors develop polarized sentiments towards individual peers but keep sentiment in balance on the ego-network level. Further, pairs of nodes tend to establish similar attitudes towards each other leading to stable and polarized positive or negative relationships.

Share

COinS
 

Sentiment Polarization and Balance among Users in Online Social Networks

Communication within online social network applications enables users to express and share sentiments electronically. Existing studies examined the existence or distribution of sentiments in online communication at a general level or in small-observed groups. Our paper extends this research by analyzing sentiment exchange within social networks from an ego-network perspective. We draw from research on social influence and social attachment to develop theories of node polarization, balance effects and sentiment mirroring within communication dyads. Our empirical analysis covers a multitude of social networks in which the sentiment valence of all messages was determined. Subsequently we studied ego-networks of focal actors (ego) and their immediate contacts. Results support our theories and indicate that actors develop polarized sentiments towards individual peers but keep sentiment in balance on the ego-network level. Further, pairs of nodes tend to establish similar attitudes towards each other leading to stable and polarized positive or negative relationships.