Discovering contextual tags from product review using semantic relatedness
Year: 2013
Editor: Udo Lindemann, Srinivasan V, Yong Se Kim, Sang Won Lee, John Clarkson, Gaetano Cascini
Author: Lim, Soon Chong Johnson; Liu, Ying
Series: ICED
Institution: 1: Universiti Tun Hussein Onn Malaysia, Malaysia; 2: National University of Singapore, Singapore
Page(s): 341-350
ISBN: 978-1-904670-49-0
ISSN: 2220-4334
Abstract
Nowadays, online product reviews has enabled product designers to better understand product related issues from the users' perspective. In the design community, there are a number of studies that have focused on studying product reviews in various analysis perspectives. While these are essential, we noticed that contextual annotation of tags has not been fully explored. We reckoned that such an annotation is equally important to better clarify the tags' context where tasks such as design experience analysis and faceted product comparison can be made possible. However, the challenge lies in automatic discovery of contextual tags from product reviews. Consequently, this paper proposed a learnable approach to address this issue. A ranking algorithm is proposed to rank important key terms along with an approach to discover contextual annotation of a given term. The performance evaluation of our proposal is done using two annotated corpus. A case study using a small laptop reviews corpus is also reported to showcase how our algorithm can be applied towards product understanding and product ontology development. Finally, we conclude this paper with some indications for future work.
Keywords: Design informatics, information management, contextual tags, semantic relatedness