Abstract: [Purpose/Significance] In the context of a rapidly aging society, smart home care for the elderly has emerged as a crucial trend. Currently, poor transparency and interpretability in human-machine interactions cause low acceptance of intelligent technology and poor trust experience among the elderly. Exploring the role of interactive control methods and agent feedback mechanisms that incorporating experiential knowledge can contribute to improve the intention of elderly users and offer a novel trajectory for the design of human-intelligent interaction. [Design/Methodology] A controlled experiment of human-intelligence interaction was constructed, and a two-factors mixed experiment of 3（interaction control mode: voice/App/mechanical）×2（interaction feedback mode: calm/active）was conducted. Three groups of subjects were recruited, with 30 older adults in each group watching two sets of interaction demonstration videos of smart cameras, and the trust experience of human-intelligence interaction was measured by scale. [Findings/Conclusion] For the privacy avoidance function of the agent, when empirical knowledge is applied to the human-intelligence interaction control method, the interaction trust experience of the aged population can be significantly affected. The trust degree of pull cord control is significantly better than that of voice control and App control and the above effect exists in the calm feedback. During active feedback, the influence of experiential knowledge on trust experience is suppressed, and there is no significant difference between pull cord control and voice control. [Originality/Value] Focusing on the influence of experiential knowledge on the trust experience of older adults in the control mode of human-intelligent interaction, this paper accurately measures the trust experience of the seniors through controlled experiments, providing a theoretical basis for optimizing the interaction experience from the level of the trust experience, and enriching the basic criteria of intelligent agents interaction design.

Keywords: Human-intelligence interaction, Trust experience, Empirical knowledge, Privacy avoidance, Older adults