Abstract
Online opinions, also referred to as customer reviews, offer a plethora of information from the customer’s perspective. By employing methods to extract the customer requirements, product designers are able to better understand the wants and needs of their customers and harness this information to meet business goals. This paper presents the process of product feature extraction and describes how this information can be used to build a Product Feature Information Hierarchy. This paper also presents a method to extract customer preference sentences from customer review data, and examines the viability of “mapping” the customer review preferences sentences to the ‘Product Feature Information Hierarchy’ using supervised learning methods. This research utilizes the My Starbucks Idea website as an online customer review site; extracting and manually reviewing over 5,100 customer reviews sentences stored on this website. The reported results provide insight into how a systematic requirement analysis can be realized using unstructured customer review data in the service product domain. Findings from this research suggest customer requirements can be extracted from unstructured text and organized in a structural way, using a Product Feature Information Hierarchy in combination with supervised learning classifiers.
Keywords
Get full access to this article
View all access options for this article.