Study of authentic and fictitious online reviews

• Main Author: Snehasish, Banerjee
• Other Authors: Alton Chua Yeow Kuan
• Format: Theses and Dissertations
• Language: English
• Published: 2017
• Subjects: DRNTU::Library and information science::Libraries::Information society; DRNTU::Business::Marketing::Consumer behavior; DRNTU::Business::Information technology::Electronic commerce; DRNTU::Library and information science::Libraries::Information retrieval and analysis
• Online Access: http://hdl.handle.net/10356/72679
• Institution: Nanyang Technological University

This research investigates the ways in which differences between authentic and fictitious reviews on the Internet are related to users’ perceived review authenticity. To accomplish the goal, it addresses two research questions. The first deals with actual review authenticity while the second focuses on humans’ perceived review authenticity. A theoretical model of authentic and fictitious reviews is developed. Guided by the model, this research examines actual differences between authentic and fictitious reviews using computational techniques. A dataset of 1,800 reviews (900 authentic + 900 fictitious) was used for investigation. Next, it examines humans’ perceived differences between authentic and fictitious reviews by surveying 380 participants. This research finds that authentic reviews are different from fictitious entries in terms of the linguistic cues of exaggeration and specificity. Moreover, humans’ perceptions of exaggeration and specificity are related to perceived authenticity. The relations tend to be particularly strong among individuals who believe that the authenticity of information cannot be justified easily. However, humans remain generally sub-par in discerning review authenticity. Overall, this research shows that even though authentic and fictitious reviews are distinguishable, human ability to discern authenticity is not overly promising. Therefore, it develops a guideline based on the actual differences between authentic and fictitious reviews to teach humans how to discern authenticity. The guideline is used as an intervention in an experimental design involving 240 participants. It is found to improve human ability to perceive differences between authentic and fictitious reviews.