Research on the interaction mechanisms of emotional healing robots moderated by attachment avoidance

doi:10.11996/JG.j.2095-302X.2026040915

Abstract

Abstract:

To address the issues of unclear cognitive mechanism and significant impact of individual psychological traits on interaction efficiency in the human-computer interaction with emotional healing robot, an emotional interaction design research was carried out for users with high and low attachment avoidance. Based on attachment theory, cognitive attention theory and human-computer interaction theory, the cognitive model of “visual attention-emotional integration” of emotional healing robots was constructed. Through a mixed experiment of 2 (anthropomorphic: high / low) ×2 (body proportion: semi-body / close-up) ×2 (interaction style: technical / social orientation), the effects of robot anthropomorphic levels, body proportion, and interaction style on emotional connection, task goal, and satisfaction of high and low attachment avoidance users were measured. Based on the eye movement experiment, the visual cognitive mechanism of high and low attachment avoidance users in early threat assessment, attention resource maintenance, and cognitive processing depth was analyzed. The SSA-BP-SHAP method was introduced to construct an emotional healing effect prediction model based on eye movement data for high and low attachment avoidance users. The study found that high and low attachment avoidance users had two different cognitive models in human-computer emotional interaction. High attachment avoidance users showed “evaluative compensation”, gaze delay when facing emotional load, and high dependence on non-core social cues such as hands to relieve stress; low attachment avoiders were characterized by “integrated immersion”, which enabled them to focus on the face to obtain emotional resonance ; there were significant differences in the influence of robot design elements on the visual attention distribution and subjective healing experience of high and low attachment avoidance users, and the prediction model could accurately quantify the relationship between visual features and healing efficacy. The research results provided a theoretical basis for the personalized design and interaction optimization of emotional healing robots for different attachment avoidance users.

Key words: emotional healing, human-robot interaction, attachment avoidance, eye-tracking, SSA-BP-SHAP

CLC Number:

TP391

LI Jie, REN Yuhang, GUO Shijie. Research on the interaction mechanisms of emotional healing robots moderated by attachment avoidance[J]. Journal of Graphics, 2026, 47(4): 915-926.

Figures/Tables 11

References 29

[1]	孙向红, 蒋毅, 陈雪峰, 等. 中国国民心理健康发展报告(2023-2024)[M]. 北京: 社会科学文献出版社, 2025: 1-25.
	SUN X H, JIANG Y, CHEN X F, et al. Report on national mental health development in China (2023-2024)[M]. Beijing: Social Sciences Academic Press (China), 2025: 1-25 (in Chinese).
[2]	TOROUS J, LINARDON J, GOLDBERG S B, et al. The evolving field of digital mental health: current evidence and implementation issues for smartphone apps, generative artificial intelligence, and virtual reality[J]. World Psychiatry, 2025, 24(2): 156-174. DOI URL
[3]	HEINZ M V, MACKIN D M, TRUDEAU B M, et al. Randomized trial of a generative AI chatbot for mental health treatment[J]. NEJM AI, 2025, 2(4): AIoa2400802.
[4]	YEN H Y, HUANG C W, CHIU H L, et al. The effect of social robots on depression and loneliness for older residents in long-term care facilities: a meta-analysis of randomized controlled trials[J]. Journal of the American Medical Directors Association, 2024, 25(6): 104979. DOI URL
[5]	ROMANO D, KARANTZAS G C, MARSHALL E M, et al. Carer burden: associations with attachment, self-efficacy, and care-seeking[J]. Journal of Social and Personal Relationships, 2022, 39(5): 1213-1236. DOI URL
[6]	MISHRA C, VERDONSCHOT R, HAGOORT P, et al. Real-time emotion generation in human-robot dialogue using large language models[J]. Frontiers in Robotics and AI, 2023, 10: 1271610. DOI URL
[7]	李洁, 张千, 沈佳骏, 等机器人拟人化形态的用户认知机理研究[J]. 机械设计与研究, 2021, 37(5): 6-9.
	LI J, ZHANG Q, SHEN J J, et al. Research on user cognition mechanism of robot anthropomorphic form[J]. Machine Design and Research, 2021, 37(5): 6-9 (in Chinese).
[8]	WANG C, HASLER S, TANNEBERG D, et al. LaMI: large language models for multi-modal human-robot interaction[C]// Extended Abstracts of the CHI Conference on Human Factors in Computing Systems. New York: ACM, 2024: 218.
[9]	ZHANG Y C, MA Y, KRAGIC D. Reframing human-robot interaction through extended reality: unlocking safer, smarter, and more empathic interactions with virtual robots and foundation models[J]. Empathic Computing, 2026, 2(2): 202533.
[10]	CHENG S W, WANG J L, TIAN J M, et al. Using humanoid robots to obtain high-quality motor imagery electroencephalogram data for better brain-computer interaction[J]. IEEE Transactions on Cognitive and Developmental Systems, 2024, 16(2): 706-719. DOI URL
[11]	BOWLBY J. Attachment and loss[M]. Harmondsworth: Penguin, 1969: 177-234.
[12]	YANG F, OSHIO A. Using attachment theory to conceptualize and measure the experiences in human-AI relationships[J]. Current Psychology, 2025, 44(11): 10658-10669. DOI
[13]	刘伟, 董悦, 李纯青. 机器人并不"冰冷": 消费者与智能社交机器人的情感关系研究综述[J]. 外国经济与管理, 2024, 46(6): 98-111.
	LIU W, DONG Y, LI C Q. Robots are not "cold": a literature review of the emotional relationship between consumers and intelligent social robots[J]. Foreign Economics & Management, 2024, 46(6): 98-111 (in Chinese).
[14]	BRENNAN K A, CLARK C L, SHAVER P R. Self-report measurement of adult attachment: an integrative overview[EB/OL]. [2026-01-15]. https://www.mendeley.com/catalogue/332b100a-3145-3597-9fa6-fba43210782c/.
[15]	李奕, 曹成才, 宋章通, 等. 机器人如何吸引儿童?外观、动作和声音多感官特征在早期互动中的作用[J]. 图学学报, 2026, 47(1): 223-233. DOI
	LI Y, CAO C C, SONG Z T, et al. How do robots attract children? The role of appearance, motion, and voice as multisensory features in early-stage interactions[J]. Journal of Graphics, 2026, 47(1): 223-233 (in Chinese). DOI
[16]	LANGNER M, TOREINI P, MAEDCHE A. Eye-based recognition of user traits and states-a systematic state-of-the-art review[J]. Journal of Eye Movement Research, 2025, 18(2): 8. DOI PMID
[17]	包子阳, 余继周, 杨杉. 智能优化算法及其MATLAB实例[M]. 3版. 北京: 电子工业出版社, 2021: 3-5.
	BAO Z Y, YU J Z, YANG S. Intelligent optimization algorithm and its MATLAB examples[M]. 3rd ed. Beijing: Publishing House of Electronics Industry, 2021: 3-5 (in Chinese).
[18]	MOHAMED Y, LEMAIGNAN S, GÜNEYSU A, et al. Fusion in context: a multimodal approach to affective state recognition[C]// The 34th IEEE International Conference on Robot and Human Interactive Communication (RO-MAN). New York: IEEE Press, 2025: 1049-1055.
[19]	纪守领, 李进锋, 杜天宇, 等. 机器学习模型可解释性方法、应用与安全研究综述[J]. 计算机研究与发展, 2019, 56(10): 2071-2096.
	JI S L, LI J F, DU T Y, et al. Survey on techniques, applications and security of machine learning interpretability[J]. Journal of Computer Research and Development, 2019, 56(10): 2071-2096 (in Chinese).
[20]	BEKLER M, YILMAZ M, ILGIN H E. Assessing feature importance in eye-tracking data within virtual reality using explainable artificial intelligence techniques[J]. Applied Sciences, 2024, 14(14): 6042. DOI URL
[21]	DE CICCO R, DA COSTA E SILVA S C L, ALPARONE F R. "It’s on its way": chatbots applied for online food delivery services, social or task-oriented interaction style?[J]. Journal of Foodservice Business Research, 2021, 24(2): 140-164. DOI URL
[22]	MAAR D, BESSON E, KEFI H. Fostering positive customer attitudes and usage intentions for scheduling services via chatbots[J]. Journal of Service Management, 2023, 34(2): 208-230. DOI URL
[23]	XIE H J, CHEN Y R, XING X F, et al. PsyDT: using LLMs to construct the digital twin of psychological counselor with personalized counseling style for psychological counseling[C]// The 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). New York: ACL, 2025: 1081-1115.
[24]	张婷, 赖建都, 侯冠华, 等. 眼控速度与目标移动距离对用户交互绩效的影响[J]. 图学学报, 2025, 46(1): 233-240. DOI
	ZHANG T, LAI J D, HOU G H, et al. The effects of eye-control speed and target travel distance on user interaction efficiency[J]. Journal of Graphics, 2025, 46(1): 233-240 (in Chinese). DOI
[25]	WEI M F, RUSSELL D W, MALLINCKRODT B, et al. The experiences in close relationship scale (ECR)-short form: reliability, validity, and factor structure[J]. Journal of Personality Assessment, 2007, 88(2): 187-204. DOI PMID
[26]	VAN ORDEN K F, LIMBERT W, MAKEIG S, et al. Eye activity correlates of workload during a visuospatial memory task[J]. Human Factors: The Journal of the Human Factors and Ergonomics Society, 2001, 43(1): 111-121. DOI URL
[27]	JAMES G, WITTEN D, HASTIE T, et al. An introduction to statistical learning: with applications in R[M]. New York: Springer, 2013: 87-97.
[28]	陈耶拉, 耿秀丽. 基于改进协同过滤的个性化产品服务系统方案推荐[J]. 计算机集成制造系统, 2021, 27(1): 240-248. DOI
	CHEN Y L, GENG X L. Recommendation of personalized product-service system scheme based on improved collaborative filtering[J]. Computer Integrated Manufacturing Systems, 2021, 27(1): 240-248 (in Chinese).
[29]	FENG Z, DANG J N. To avoid disconnection or approach connection: loneliness predicts social robot anthropomorphism via different social motivations in the UK and China[J]. Personality and Individual Differences, 2025, 235: 112979. DOI URL

维度	题项
情感联结	在交互过程中感受到了真诚在交互过程中感受到了被理解在交互过程中感受到了亲切与放松
任务目标	在交互过程中感受到了易于交谈性在交互过程中感受到了专业性在交互过程中感受到了可靠性
满意度	这段交互体验是愉快的，以后我愿意继续与它交流

维度	题项
情感联结	在交互过程中感受到了真诚在交互过程中感受到了被理解在交互过程中感受到了亲切与放松
任务目标	在交互过程中感受到了易于交谈性在交互过程中感受到了专业性在交互过程中感受到了可靠性
满意度	这段交互体验是愉快的，以后我愿意继续与它交流

回避分组	首次注视时间/s		注视总持续时间/s		总访问持续时间占比		平均注视持续时间/s
回避分组	高	低	高	低	高	低	高	低
截距	145.35***	160.43***	508.34***	610.82***	1010.66***	1601.75***	342.34***	362.76***
拟人化	13.06**	4.83*	49.50***	43.43***	29.40***	39.51***	12.66**	11.03***
交互风格	0.06	0.20	0.01	0.13	0.82	0.33	0.99	0.12
身体比例	13.67***	50.46***	18.07***	29.41***	76.58***	118.30***	1.17	5.62*
拟人化×交互风格	0.69	5.26*	0.62	0.00	0.03	0.53	0.73	0.39
拟人化×身体比例	1.25	7.37**	0.30	0.56	1.56	7.83**	1.81	0.22
交互风格×身体比例	0.40	0.00	0.00	0.18	0.11	0.62	0.06	0.79
拟人化×交互风格×身体比例	0.11	0.01	0.00	0.17	0.68	0.01	0.45	1.07

回避分组	首次注视时间/s		注视总持续时间/s		总访问持续时间占比		平均注视持续时间/s
回避分组	高	低	高	低	高	低	高	低
截距	145.35***	160.43***	508.34***	610.82***	1010.66***	1601.75***	342.34***	362.76***
拟人化	13.06**	4.83*	49.50***	43.43***	29.40***	39.51***	12.66**	11.03***
交互风格	0.06	0.20	0.01	0.13	0.82	0.33	0.99	0.12
身体比例	13.67***	50.46***	18.07***	29.41***	76.58***	118.30***	1.17	5.62*
拟人化×交互风格	0.69	5.26*	0.62	0.00	0.03	0.53	0.73	0.39
拟人化×身体比例	1.25	7.37**	0.30	0.56	1.56	7.83**	1.81	0.22
交互风格×身体比例	0.40	0.00	0.00	0.18	0.11	0.62	0.06	0.79
拟人化×交互风格×身体比例	0.11	0.01	0.00	0.17	0.68	0.01	0.45	1.07

回避分组	首次注视时间/s		注视总持续时间/s		总访问持续时间占比		平均注视持续时间/s
回避分组	高	低	高	低	高	低	高	低
截距	110.49***	117.04***	138.90***	95.35***	136.05***	< 0.001	287.36***	303.64***
拟人化	26.71***	34.80***	2.16	0.98	0.01	0.382	5.04*	0.88
交互风格	1.19	2.01	6.61*	0.01	6.76*	0.785	0.39	1.15
拟人化×交互风格	0.02	5.19*	4.99*	0.54	5.06*	0.946	0.40	4.26*