Interaction Design in Human-in-the-Loop Hybrid Intelligence
Yuanda Hu, Xiaohua Sun, Hexin Zhang, Shengchen Zhang, Siqi Yi
Abstract
The work aims to point out the problems in “interaction design for human-in-the-loop hybrid intelligent” and study the problems of interaction design in the human-in-the-loop hybrid intelligent system, to provide references for the related research in design, technology and application. Starting from the concept and architecture of human-in-the-loop hybrid intelligence, the interaction design of human-in-the-loop hybrid intelligence was introduced and the interface compositions and interaction models and life cycle of human-in-the-loop hybrid intelligence were summarized based on related work in different fields. It is pointed out that the human-in-the-loop hybrid intelligence requires the intelligent model of user interaction and the system architecture composed of users, artificial intelligence algorithms, and user interfaces is introduced. The interaction methods of existing work for different data types are summarized, the design challenges for each key activity of the full life cycle in human-in-loop hybrid intelligence are analyzed, the characteristics and key interface composition of different user functions and requirements are extracted, and the design principles of the human-in-the-loop hybrid intelligent system are proposed. The suggestions for establishing design methodology from intelligent system, user and designer are proposed, improving design tools to support and promote the applications of human-in-the-loop hybrid intelligent systems.