An encounter-based social link prediction method using Bluetooth traces
Thi-Nga Dao, Seokhoon Yoon · 2017
This work considers the social link prediction problem in the human community. Since mobile phones with the support of wireless technologies have become an integral part of human life, mining the traces of mobile devices can help us to understand the community structure in the social network. Therefore, an encounter-based link prediction method in which we try to analyze the social relationship between subjects using Bluetooth logs is proposed. In order to evaluate the obtained social ties, a user position prediction model, that applies two machine learning techniques, with the support of the location information of strongest extracted social friends is used.