Understanding social relationship evolution by using real-world sensing data

Mobile and pervasive computing technologies enable us to obtain real-world sensing data for sociological studies, such as exploring human behaviors and relationships. In this paper, we present a study of understanding social relationship evolution by using real-life anonymized mobile phone data. First, we define a friendship as a directed relation, i.e., person A regards another person B as his or her friend but not necessarily vice versa. Second, we recognize human friendship from a supervised learning perspective. The Support Vector Machine (SVM) approach is adopted as the inference model to predict friendship based on a variety of features extracted from the mobile phone data, including proximity, outgoing calls, outgoing text messages, incoming calls, and incoming text messages. Third, we demonstrate the social relation evolution process by using the social balance theory. For the friendship prediction, we achieved an overall recognition rate of 97.0 % by number and a class average accuracy of 89.8 %. This shows that social relationships (not only reciprocal friends and non-friends, but non-reciprocal friends) can be likely predicted by using real-world sensing data. With respect to the friendship evolution, we verified that the principles of reciprocality and transitivity play an important role in social relation evolution.