Association rule mining and cognitive pairwise rating based portfolio analysis for product family design

Abstract

Changing customer needs coupled with rapid technology advances has boosted stronger requirements for a greater variety of consumer electronics. This trend has forced global companies to reconsider their product-positioning strategies. To reduce design cost and shorten the time to market, portfolio analysis for product family design is usually adopted to acquire diverse but related market applications. This study presents a novel framework to implement product differentiation and product configuration. Initially, association rule mining is used to capture user perceptions to identify the significant portfolios of hedonic attributes. Secondly, cognitive pairwise rating is conducted to elicit user preferences for utilitarian attributes (UAs). Finally, the Technique for Order Preference by Similarity to Ideal Solution is used to prioritize the optimal portfolios of UAs. Experiment results show that “keyboard interface”, “body material”, and “screen size” are the most concerned HAs for differentiating the product family while “CPU performance” is the most important UA for configuring padbooks, ultrabooks and notebooks. In summary, this research allows companies to effectively and efficiently incorporate user perceptions or preferences into the entire decision-making process.