Dialogue systems must understand children’s utterance intentions by considering their unique linguistic characteristics, such as syntactic incompleteness, pronunciation inaccuracies, and creative expressions, to enable natural conversational engagement in child–robot interactions. Even state-of-the-art large language models (LLMs) for language understanding and contextual awareness cannot comprehend children’s intent as accurately as humans because of their distinctive features. An LLM-based dialogue system should acquire the manner by which humans understand children’s speech to enhance its intention reasoning performance in verbal interactions with children. To this end, we propose a fine-tuning methodology that utilizes the LLM–human judgment discrepancy and interactive response data. The former data represent cases in which the LLM and human judgments of the contextual appropriateness of a child’s answer to a robot’s question diverge. The latter data involve robot responses suitable for children’s utterance intentions, generated by the LLM. We developed a fine-tuned dialogue system using these datasets to achieve human-like interpretations of children’s utterances and to respond adaptively. Our system was evaluated through human assessment using the Robotic Social Attributes Scale (RoSAS) and Sensibleness and Specificity Average (SSA) metrics. Consequently, it supports the effective interpretation of children’s utterance intentions and enables natural verbal interactions, even in cases with syntactic incompleteness and mispronunciations.

This study proposes a dual-level analytical approach to observing human–robot interactions in a real-world public setting, specifically a science museum. Observation plays a crucial role in human–robot interaction research by enabling the capture of nuanced and context-sensitive behaviors that are often missed by post-interaction surveys or controlled laboratory experiments. Public environments such as museums pose particular challenges due to their dynamic and open-ended nature, requiring methodological approaches that balance ecological validity with analytical rigor. To address these challenges, we introduce a dual-level approach for behavioral observation, integrating statistical analysis across demographic groups with time-series modeling of individual engagement dynamics. At the group level, we analyzed engagement patterns based on age and gender, revealing significantly higher interaction levels among children and adolescents compared to adults. At the individual level, we employed temporal behavioral analysis using a Hidden Markov Model to identify sequential engagement states—low, moderate, and high—derived from time-series behavioral patterns. This approach offers both broad and detailed insights into visitor engagement, providing actionable implications for designing adaptive and socially engaging robot behaviors in complex public environments. Furthermore, it can facilitate the analysis of social robot interactions in everyday contexts and contribute to building a practical foundation for their implementation in real-world settings.

This study proposes a dual-level analytical approach to observing human-robot interactions in a real-world public setting, specifically a science museum. Observation plays a crucial role in human-robot interaction research by enabling the capture of nuanced and context-sensitive behaviors that are often missed by post-interaction surveys or controlled laboratory experiments. Public environments such as museums pose particular challenges due to their dynamic and open-ended nature, requiring methodological approaches that balance ecological validity with analytical rigor. To address these challenges, we introduce a dual-level approach for behavioral observation, integrating statistical analysis across demographic groups with time-series modeling of individual engagement dynamics. At the group level, we analyzed engagement patterns based on age and gender, revealing significantly higher interaction levels among children and adolescents compared to adults. At the individual level, we employed temporal behavioral analysis using a Hidden Markov Model to identify sequential engagement states—low, moderate, and high—derived from time-series behavioral patterns. This approach offers both broad and detailed insights into visitor engagement, providing actionable implications for designing adaptive and socially engaging robot behaviors in complex public environments. Furthermore, it can facilitate the analysis of social robot interactions in everyday contexts and contribute to building a practical foundation for their implementation in real-world settings.