Home // eKNOW 2018, The Tenth International Conference on Information, Process, and Knowledge Management // View article


Everything Was Good! The Influence of Sentiment and Product Category on Aspect Choice in German Customer Reviews

Authors:
Amelie I. Metzmacher
Verena Heinrichs
Björn Falk
Robert H. Schmitt

Keywords: NLP; sentiment analysis; social media

Abstract:
Studies dealing with the use of pronouns in customer reviews in social media against the background of sentiment analysis do not consider the role attributed to the pronoun within an evaluative sentence. The following paper addresses the question of when and why customers choose pronouns instead of nouns or proper nouns to function as aspects within an evaluative sentence. To this end, two influencing factors on aspect choice have been investigated: 1) The sentiment of the evaluative sentence (positive, neutral, negative); 2) The category of the evaluated aspect (product, service). The results reveal that, particularly within positive statements, there exist differences between customer reviews evaluating products to those evaluating services.

Pages: 32 to 35

Copyright: Copyright (c) IARIA, 2018

Publication date: March 25, 2018

Published in: conference

ISSN: 2308-4375

ISBN: 978-1-61208-620-0

Location: Rome, Italy

Dates: from March 25, 2018 to March 29, 2018