Emotions Embedded in Online Reviews and Social Influence: an Abstract

EasyChair Preprint 10652

Abstract

Consumers are increasingly relying on the aggregate opinion of others to make purchasing decisions by reading online reviews. Existing research has shown that previously posted ratings cause significant social influence bias in individual rating behavior, but there has been little research into how social influence impact on subsequent ratings varies with emotions embedded in reviews, product, and reviewer characteristics. The authors used a huge dataset—over 150 thousand online reviews from TripAdvisor to extract emotions embedded in reviews using the newest text mining technique, including multiple machine learning algorithms, to examine the moderators of social influence impact on subsequent ratings. The results show that social influence have a stronger influence on subsequent ratings when the customer has a negative experience and the emotion expressed in reviews is anger, whereas the influence is weaker when the customer has an extreme positive experience and the emotion expressed in a review is joy.

Keyphrases: Emotions, Ratings, Text Analytics, social influence

@booklet{EasyChair:10652,
  author    = {Feray Adiguzel and Moamen Elsherbiny and Carmela Donato and Evangelos Syrigos},
  title     = {Emotions Embedded in Online Reviews and Social Influence: an Abstract},
  howpublished = {EasyChair Preprint 10652},
  year      = {EasyChair, 2023}}