An innovative data mining-driven optimisation modelling approach based on TAM for the design of elderly-centric ICT products

To enhance the quality of life and health of the elderly, it is necessary to incorporate their specific needs into the design and manufacturing processes of Information and Communication Technology (ICT) products. However, these needs are often overlooked in product development, leading to poor user experiences and low adoption rates. To address this important issue, this study innovatively proposes a data mining-driven optimisation modelling approach. This approach first extracts key design elements from online reviews of elderly ICT products through text mining techniques such as data purification and Latent Dirichlet Allocation (LDA). These elements are then incorporated as external variables to construct a user needs integrated technology acceptance model (UN-TAM) for the elderly, which is validated using Structural Equation Modelling (SEM). Using the elderly smartwatch as a case study, four design optimisation parameters, functional architecture, human-computer interface, morphological aesthetics, and interaction mode, were extracted from a substantial database based on the elderly user comments. The proposed model establishes a quantitative mapping relationship among “design parameters, user perception, and behavioral intention”. The results indicate that these design parameters significantly influence the product acceptance through three distinct pathways: perceived ease of use (γ = 0.641, p < 0.001), perceived usefulness (γ = 0.601, p < 0.001), and perceived enjoyment (γ = 0.402, p < 0.001). Specifically, functional architecture of the elderly-centric ICT products has the most substantial impact on perceived usefulness while the other three parameters predominantly influence perceived ease of use. The proposed data mining-driven optimisation modelling approach can be effectively used in designing and manufacturing elderly-centric ICT products.