Host: The Japanese Society for Artificial Intelligence
Name : The 38th Annual Conference of the Japanese Society for Artificial Intelligence
Number : 38
Location : [in Japanese]
Date : May 28, 2024 - May 31, 2024
Conversations between people are established through mutual understanding. In such conversations, there are cases in which people complement each other's utterances or overlap each other's utterances. Such communication is called co-speech. In this study, we investigated whether co-speech is a communication style in which common ground is easily formed and renewed. We conducted a joint figure placement task and a keyword generation task, and examined whether there is a difference in the frequency of collocations between pairs with more common ground and other groups, and examined the relationship between collocations and the formation and renewal of common ground. The results suggest that pairs with a high frequency of co-speech may be more likely to agree on the placement of figures, and that paraphrase-type co-speech is often observed in the formation stage of the common ground. We hope that this research will lead to the diversification of communication with robots.