Research On Facial Emotion Expression Design Of Educational Robot Based On Conjoint Analysis Method

With the development of artificial intelligence technology and the prosperity of service robots,educational robots have entered a new stage of development.They not only have a substantial knowledge base and intelligent interactive technology,but also can express rich emotions,providing students with a good learning experience and a more natural way of communication.This paper discusses the research status,basic connotation and research trend of educational robot through literature analysis.The concept of robot emotion,the way of robot emotion expression and the factors affecting robot emotion expression are also discussed.It is proposed that the research on the emotional expression of educational robots is an important development direction in the future and has important research significance.Designers make robots have the ability of emotional expression by imitating human or animal emotions.Therefore,it is necessary to analyze the intrinsic relationship between emotional expression elements and characteristics of educational robots and specific emotions,so as to realize more real and effective emotional interaction experience and achieve corresponding teaching objectives.This paper aims to explore the importance weights and utility values of educational robots’ facial features for emotional expression by using the conjoint analysis method.Then proposes the full contour utility model and obtains the optimal utility combination for facial emotional expression in the conjoint analysis method.Firstly,the categories of emotions（happiness and sadness）were established through literature analysis.Secondly,8 elements and 23 features of facial expressions were selected,and a questionnaire was designed according to the seven-point semantic scale.Finally,the importance weights and utility values of facial emotional expression of educational robots were obtained by conjoint analysis method.Combined with the full contour utility model,a specific design strategy is proposed from four perspectives,namely,the unity of design,the difference of emotion expression,the limitation of objective conditions and the development of the industry.The proposed design method is used to guide the optimization of facial expression design,which provides an effective reference for the design of facial expression of educational robots.By using behavioral and performance measures of distraction times,error rate and quiz time,this paper explores the influence of the optimal utility combination of emotional expression on the learning effect of educational robot,and analyzes the correlation between such influence and gender and task difficulty,and verifies the effectiveness of the optimal utility combination from the side.