Title: Review rating prediction using explicit and implicit features

Authors: Debasmita Dey; Pradeep Kumar

Addresses: Indian Institute of Management Lucknow, UP, 226013, India ' Indian Institute of Management Lucknow, UP, 226013, India

Abstract: In e-commerce websites, customers can rate the quality of the product on a 1-5 scale rating along with a detailed feature-wise review of it. Due to the availability of the same product in multiple e-commerce websites and variability in the number of customers in each e-commerce platform, the number of ratings obtained differs across various sites. Hence, it is observed that sometimes the overall rating (an indicator of quality) of the same product differs significantly across multiple e-commerce platforms, which creates anxiety and distrust in customers' purchasing decisions and may cause loss of an online sale. In this present study, we propose a model, which produces a single score of any product, and this score is more reliable and trustworthy than the conventional method as it captures both explicit and implicit information instead of considering just either of them.

Keywords: e-commerce; online reviews; ratings; prediction.

DOI: 10.1504/IJBIS.2025.149539

International Journal of Business Information Systems, 2025 Vol.50 No.3, pp.333 - 351

Received: 06 Jul 2021
Accepted: 30 Nov 2021
Published online: 07 Nov 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article