Analysis of body motion synchrony phenomenon in communities and between communities

We evaluate community activity in social network based on body motion synchrony of two people during face-to- face communication. In particular, we look at people's body motion synchrony when they are in the same communities and from different communities. Using wearable sensors, we measured individuals' time series body motion data and face-to- face communication data. From these data we detected communities in 6 organizations and statistically analyze the distribution of body motion rhythm difference in communities and between communities. The result showed the tendency that people who are in the same communities are easier to synchrony than people who are from different communities. Moreover, we make comparison on the result based on two different community detection methods. One detection method is based on real department information, the other one is based on real interaction information. The result showed that the above tendency is more common in community separation based on real interaction information. The present study will create a new path to evaluate communities detected in different community detection methods in terms of body motion synchrony.