Extraction and Design of Favorite Products Through Analyzing Customer Latent Preferences

Ryosui Koga, Hideki Aoyama

Research output: Contribution to journalArticlepeer-review


In the wake of rapid advances in design and production technologies, differentiating products based on their quality has become a challenge. Against this backdrop, design has become an important factor in determining product value. Design is a creative activity influenced by the experience and sensitivity of designers, who are required to understand the preferences and needs of customers and reflect them in their designs. Accordingly, there is a need to efficiently determine customer preferences. Although it is possible to extract customers’ apparent preferences through interviews and questionnaires, these may be arbitrary. Additionally, to respond to the recent diversification of customer preferences, it is not enough to understand apparent preferences; latent preferences must also be extracted. However, they are vague and cannot be expressed in words by the customers. Unfortunately, a practical method for extracting latent preferences has not yet been developed. In this study, we propose a method for extracting latent customer preferences. We develop a system for recommending products that customers are likely to prefer from among existing products, and develop a system for creating original product designs that customers are expected to prefer. We experimentally verify the usefulness of this method.

Original languageEnglish
Pages (from-to)807-813
Number of pages7
JournalInternational Journal of Automation Technology
Issue number6
Publication statusPublished - 2022 Nov


  • behavior observation
  • latent preference
  • neural network
  • product design

ASJC Scopus subject areas

  • Mechanical Engineering
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'Extraction and Design of Favorite Products Through Analyzing Customer Latent Preferences'. Together they form a unique fingerprint.

Cite this