Incorporating user preferences and performance ratings into multi-functional tablet design and recommendation

Chih Hsuan Wang*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


The boundaries between smartphones, tablets, and ultrabooks are becoming blurred nowadays. Inspired by the concept of product family design, a novel framework is presented to help brand companies generate and assess innovative ideas to construct a recommender system for designing multi-functional tablets. First, the theory of inventive problem solving (TRIZ) is used to tackle the trade-offs between the improved features (area of a stationary object and ease of operation) and the worsening features (area, length, and weight of a stationary object). Rather than limiting to compromised solutions, innovative ideas are generated for handling engineering conflicts. In particular, four inventive principles are identified to be feasible solutions: #1 (segmentation) is for designing PadBook (tablet and ultrabook), #7 (nesting) and #30 (flexible membranes or thin film) is for designing PadFone (tablet and smartphone), and #15 (dynamicity) is for designing PadNote (tablet with a keyboard and stylus pen). To achieve product recommendation in an unsupervised way, fuzzy analytical hierarchy process (FAHP) is employed to extract user preferences for product features and user ratings of product families. Finally, classification and regression tree (CART) is constructed to justify product recommendation in a supervised way. In summary, the presented framework can help product planners generate innovative ideas, capture user preferences, aggregate performance ratings, and construct a decision-support system for accomplishing innovative tablet design and product recommendation.

Original languageEnglish
Pages (from-to)308-320
Number of pages13
JournalInternational Journal of Industrial Engineering : Theory Applications and Practice
Issue number2
StatePublished - 2020


  • CART
  • FAHP
  • Multi-functional tablets
  • Recommender system
  • TRIZ

Fingerprint Dive into the research topics of 'Incorporating user preferences and performance ratings into multi-functional tablet design and recommendation'. Together they form a unique fingerprint.

Cite this