Start Date
16-8-2018 12:00 AM
Description
As online reviews provide essential information to guide customers in their prospective purchases. As more such reviews accumulate overtime, one would suspect that complete information becomes available about product and almost no customers should be disappointed in their purchases. Yet, we provide empirical evidence over a large dataset of reviews that negative reviews seem to be arriving at an accelerated rate later on in the lifetime of a product. To better understand this inconsistency, we frame the problem at hand as an information revelation problems. Using a novel approach, we then segment each review as an aggregation of aspects for which the reviewer provides weights, in line with how much she values those aspects, and corresponding experiences vis-Ã -vis those aspects, which range from positive to negative. We show that this segmentation better explains the review process and better explains the polarity of reviews.
Recommended Citation
Cheng, Yichen; Jabr, Wael; Srivastava, Sanjay; and Zhao, Kai, "Dynamics of Information Revelation in Online Reviews" (2018). AMCIS 2018 Proceedings. 21.
https://aisel.aisnet.org/amcis2018/eBusiness/Presentations/21
Dynamics of Information Revelation in Online Reviews
As online reviews provide essential information to guide customers in their prospective purchases. As more such reviews accumulate overtime, one would suspect that complete information becomes available about product and almost no customers should be disappointed in their purchases. Yet, we provide empirical evidence over a large dataset of reviews that negative reviews seem to be arriving at an accelerated rate later on in the lifetime of a product. To better understand this inconsistency, we frame the problem at hand as an information revelation problems. Using a novel approach, we then segment each review as an aggregation of aspects for which the reviewer provides weights, in line with how much she values those aspects, and corresponding experiences vis-Ã -vis those aspects, which range from positive to negative. We show that this segmentation better explains the review process and better explains the polarity of reviews.