Designing Personalization Cues for Museum Robots: Docent Observation and Controlled Studies.

Area of Science:

• Human-Computer Interaction
• Robotics
• Social Psychology

Background:

• Public cultural venues like science museums require social robots to engage diverse visitors in short, one-off interactions.
• Long-term user modeling is impractical in these settings, necessitating alternative personalization strategies.

Purpose of the Study:

• To identify immediately interpretable behavioral cues for social robots that can evoke personalization without user profiling.
• To investigate how specific robot behaviors influence user perception of personalization and intelligence.

Main Methods:

• Observational study of museum docents to identify effective social interaction strategies.
• Three controlled laboratory studies using video-based and Wizard-of-Oz (WoZ) methods to test robot cues.
• Examined recognition accuracy, knowledge alignment, preference inquiry, and memory-based continuity.

Main Results:

• Robot recognition accuracy positively influenced social impressions of intelligence.
• Knowledge alignment, explicit preference inquiry, and memory-based continuity cues significantly increased perceived personalization.
• Micro-level personalization cues are effective in short-term encounters.

Conclusions:

• Social robots can achieve personalization in public venues through interpretable, short-term behavioral cues.
• Findings support user-centered design for social robots in public environments, enhancing visitor engagement.
• Effective personalization does not require extensive user data or long-term profiling.