A Fast and Accurate Approach for Inferencing Social Relationships Among IoT Objects

The Internet of Things (IoT) has recently moved towards the "object-object" interaction model where things look for other things to provide composite services for the benefit of human beings, leading to the birth of the Social Internet of Things (SIoT) paradigm. Investigating the social dimension in IoT objects offers great opportunities to increase social awareness among IoT objects. To achieve this goal, recurrent spatio-temporal meetings among IoT objects could be exploited to enable smart objects to understand the co-presence with other smart objects. Therefore, we target to explore the social dimension by determining if any two IoT objects have met at a particular place for a period of time. In this paper, we develop a novel approach, named Social Relationships Inference (SociRence), based on computational geometry to calculate the co-presence among IoT objects efficiently. We conduct experimental studies on real-world SIoT datasets to evaluate the efficacy of our approach. The results demonstrate that our approach can calculate the spatio-temporal co-presence at a much higher speed than the baseline computation methods.