Analyzing and visualizing comprehensive and personalized online product reviews

초록

This paper presents a method of analyzing and visualizing e-commerce sites’ product reviews considering customers’ characteristics. Typically, product review data are unstructured and have no fixed format. Such data can be considered as valuable contents or assets generated by customers. Indeed, they can perform significant roles for customers’ benefit. They provide potential consumers with helpful product information (e.g., characteristics) according to which they can decide whether to buy or not. Also, e-commerce companies can understand customers’ experiences or opinions on a product and reflect them when developing marketing strategies or redesigning the product. In order to provide valuable information to customers from massive and unstructured review data, we propose a comprehensive framework that encompasses the processes of collecting, storing, preprocessing and analyzing review data and, finally, deriving implications therefrom. Text mining and personalization techniques can be applied to the process of extracting and visualizing the review data, as appropriate. Thus, the proposed system, as it considers customer profiles when analyzing and visualizing product reviews, enables customers to utilize review data conveniently in accordance with their interests. © 2018 Springer Science+Business Media, LLC, part of Springer Nature

키워드