Explore And Classify User Generated Product Reviews From The Internet: A Case Study
Theron Pieterse, I. Perez Artieda, Aylin Koca, Yuan Lü · TU/e Research Portal · 2010
The overall aim of this paper is to demonstrate how to obtain reliable information from the user generated product reviews on the Internet. Most of these reviews only contain information about products’ soft failures, which result in the rejection of innovative consumer electronics products that are not due to broken hardware or software, but due to "broken expectations" of users. However, the reliability of this essentially subjective information is difficult to evaluate. In order to evaluate the reliability of the user generated product reviews, we explore a method to first classify them using the Disconfirmed Expectations Ontology (DEO) that is developed to analyze soft reliability issues, and then further to explore their information content by means of user tests. We apply this method in a case study using a consumer electronic product from a multi-national company.