Notes
Overcoming communicative ambiguity to strengthen the bond between humans and robots
329Social relationships between humans and social robots are crucial for effective communication and collaboration. The current project studies how social communication and relationships between robots and humans develop as two sides of the same coin: bonding results in better communication and better communication enhances affective bonding. Integrating both perspectives, our theoretical framework combines the Theory of Affective Bonding from a media-psychological perspective (emphasizing human processing) with the Theory of Identity, Reference, and Perspective from a linguistics perspective (e.g., overcoming ambiguity) to optimally program social robots for communicative purposes. For example, the same word can have different meanings (e.g., bank) and different words are used to refer to the same (e.g., flat, apartment). Likewise, humans use various non-verbal cues to coordinate conversation (e.g., eye blinking as turn-taking cue). The research question is how to achieve the natural feel of human conversation through fabricated non-human cues to develop autonomously functioning robots for human-robot communication.
330We examined how participants (both children and adults) develop communication and bonding over time through playing a game together at festivals. The game has been developed such that the robot and participant need to find common ground in searching for differences between pictures and increase their mutual understanding (cf. ‘Spot the difference’, a shared memory game). To test assumptions, we applied a between-subjects repeated measures experimental design in which communicative cues were varied through two factors (Factor 1: Rational Speech Act Model vs. ChatGPT; Factor 2: fabricated explicit turn-taking cues vs. intuitive implicit turn-taking). Participants played the game with the robot over 6 different sessions. Observational audio-visual data were collected during gameplay, whereas questionnaires (e.g., dimensions of bonding, understanding) were taken beforehand, after the first trial, and afterwards. This long-term design allows to study mutual language understanding and human-robot bonding over time, which is unique in the field.
331Human-robot communication / computational linguistics / mutual understanding / affective bonding / field experiment