[Paper Review] On the impact of robot personalization on human-robot interaction: A review

Jinyu Yang, Camille Vindolet+2 more|arXiv (Cornell University)|Jan 22, 2024

Social Robot Interaction and HRIPsychology3 citations

TL;DR

This review examines the impact of robot personalization on human-robot interaction (HRI), analyzing strategies, personalized features, technologies, and use cases across healthcare, education, and elderly care. It finds that personalization significantly enhances user engagement, trust, and learning outcomes, but highlights a critical research gap in understanding its potential negative effects on privacy, ethics, and autonomy.

ABSTRACT

This study reviews the impact of personalization on human-robot interaction. Firstly, the various strategies used to achieve personalization are briefly described. Secondly, the effects of personalization known to date are discussed. They are presented along with the personalized parameters, personalized features, used technology, and use case they relate to. It is observed that various positive effects have been discussed in the literature while possible negative effects seem to require further investigation.

Motivation & Objective

• To systematically review existing strategies for personalizing robots in human-robot interaction (HRI).
• To analyze the effects of personalization on user experience, trust, and interaction quality across diverse use cases.
• To identify gaps in current research, particularly regarding potential negative impacts such as privacy violations and ethical risks.
• To examine the distinction between perceived and actual personalization and its implications for user perception and trust.
• To guide future research by highlighting under-investigated areas, especially ethical and security concerns in personalized HRI.

Proposed method

• Conducted a comprehensive literature review of studies on robot personalization in HRI from 2012 to 2023.
• Categorized personalization strategies based on personalized parameters (e.g., behavior, emotional state, preferences), features (e.g., social interaction, feedback), and technologies (e.g., ROS, BKT models, sensors).
• Mapped personalization approaches to specific use cases, including education, elderly care, and social companionship.
• Evaluated the impact of personalization using reported outcomes such as engagement, learning gains, and user trust metrics.
• Identified methodological and ethical concerns, including perceived vs. actual personalization and risks of emotional manipulation.
• Synthesized findings to highlight both positive outcomes and under-researched risks, particularly in privacy, trust, and ethical design.

Figure 1. An overview of the personalized topics proposed by Min Kyung Lee et al. (Lee et al . , 2012 )

Experimental results

Research questions

• RQ1 How do different personalization strategies influence user engagement and trust in human-robot interaction?
• RQ2 What are the key technological and behavioral components enabling effective robot personalization across diverse applications?
• RQ3 To what extent does perceived personalization differ from actual personalization, and how does this affect user perception?
• RQ4 What are the potential negative impacts of robot personalization on user privacy, autonomy, and ethical integrity?
• RQ5 How can personalization be designed to balance user experience with ethical considerations such as fairness and transparency?

Key findings

• Personalization significantly increases children’s engagement and valence during second language tutoring, with one study reporting a notable rise in positive emotional states.
• Robots that adapt to children’s behavior—such as adjusting challenge levels or responding to help requests—demonstrate improved learning outcomes and longer-term engagement, particularly in children with autism spectrum disorder (ASD).
• Personalized curriculum sequencing based on real-time performance data led to measurable learning gains in English language learning tasks.
• Perceived personalization often matters more than actual personalization; users responded more positively to messages they believed were personalized, even when they were not.
• Personalization strategies that incorporate emotional and behavioral feedback (e.g., via electrodermal activity or facial recognition) enhance user impressions and perceived trustworthiness.
• Despite strong positive effects, the literature shows a significant lack of research on negative consequences, including privacy breaches, emotional manipulation, and erosion of user autonomy.