User experience research on intimacy and usability when an ai voice-activated personal assistant uses a dialect

Abstract

Background Personification is becoming important in efforts to increase the intimacy between users and artificial intelligence assistants. While previous studies have tried to personify artificial intelligent assistants in various ways, there has been no research into how user intimacy changes when a voice-activated personal assistant (VAPA) uses a dialect. Methods One hundred participants from Seoul (25), Chungcheongdo (25), Gyeongsangdo (25), and Jeonrado (25) were recruited through convenience sampling. Every participant was asked to try four types of VAPA (standard language, Chungcheongdo dialect, Gyeongsangdo dialect, and Jeonrado dialect) and randomly told four of 16 pre-written commands. After testing the VAPA, each participant was asked to complete three questionnaires (emotional intimacy, cognitive intimacy, and system usability scale questionnaires). Results In terms of emotional intimacy, participants from all regions except Seoul felt a higher level of intimacy when using VAPA with a dialect from the participant's region. In terms of cognitive intimacy, every participant felt more intimate with a VAPA that used the dialect native to their region. In terms of usability, all participants rated the VAPA that used the dialect from their region as more usable than a VAPA that used a dialect from a different region. However, there was no difference in usability between VAPA that used standard language and a VAPA that used a dialect from the participant's region. Conclusions The results of this paper could help companies develop detailed strategies to increase the level of intimacy between artificial intelligence agents and users.