Voice of the customers: mining online customer reviews for product feature-based ranking

  • Authors:

  • Kunpeng Zhang;Ramanathan Narayanan;Alok Choudhary

  • Affiliations:

  • Electrical Engineering and Computer Science Department, Northwestern University, Evanston, IL;Electrical Engineering and Computer Science Department, Northwestern University, Evanston, IL;Electrical Engineering and Computer Science Department, Northwestern University, Evanston, IL

  • Venue:
  • WOSN'10 Proceedings of the 3rd conference on Online social networks
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Increasingly large numbers of customers are choosing online shopping because of its convenience, reliability, and cost. As the number of products being sold online increases, it is becoming increasingly difficult for customers to make purchasing decisions based on only pictures and short product descriptions. On the other hand, customer reviews, particularly the text describing the features, comparisons and experiences of using a particular product provide a rich source of information to compare products and make purchasing decisions. Online retailers like Amazon.com1 allow customers to add reviews of products they have purchased. These reviews have become a diverse and reliable source to aid other customers. Traditionally, many customers have used expert rankings which rate limited a number of products. Existing automated ranking mechanisms typically rank products based on their overall quality. However, a product usually has multiple product features, each of which plays a different role. Different customers may be interested in different features of a product, and their preferences may vary accordingly. In this paper, we present a feature-based product ranking technique that mines thousands of customer reviews. We first identify product features within a product category and analyze their frequencies and relative usage. For each feature, we identify subjective and comparative sentences in reviews. We then assign sentiment orientations to these sentences. By using the information obtained from customer reviews, we model the relationships among products by constructing a weighted and directed graph. We then mine this graph to determine the relative quality of products. Experiments on Digital Camera and Television reviews from real-world data on Amazon.com are presented to demonstrate the results of the proposed techniques.