Title: Design feature opinion cause analysis: a method for extracting design intelligence from web reviews
Authors: Ismail Art Yagci; Sanchoy Das
Addresses: Department of Mechanical and Industrial Engineering, New Jersey Institute of Technology, University Heights, Newark, NJ 07102, USA ' Department of Mechanical and Industrial Engineering, New Jersey Institute of Technology, University Heights, Newark, NJ 07102, USA
Abstract: Web reviews are a readily available source of product intelligence. The hypothesis of this research is that web reviews also contain significant amounts of design information for product designers. The authors introduce the design-feature-opinion-cause relationship (DFOC) method, which effectively extracts design intelligence from unstructured web reviews. The DFOC method: 1) creates a sentence-based web review database; 2) mines the database to identify design features that are of interest to both designers and users; 3) extracts and estimates the significance and polarity of the customer opinions; 4) identifies the likely design cause of the customer opinion. DFOC utilises an association rule-based opinion mining procedure for capturing and extracting noun-adjective and noun-verb relationships. Application of the DFOC method utilising RapidMiner is demonstrated for an automobile. Example features and feature-opinion-cause associations are shown along with the observed opinion polarity score and cause association strength.
Keywords: DFOC; opinion mining; web reviews; design features; feature extraction; feature opinion; sentiment analysis; design intelligence; product intelligence; product design; customer opinions; association rules mining; vehicle design.
International Journal of Knowledge and Web Intelligence, 2015 Vol.5 No.2, pp.127 - 145
Available online: 05 Mar 2016 *Full-text access for editors Access for subscribers Purchase this article Comment on this article