Social signature identification of dynamical social networks

Understanding the social relation of dynamical online social networks (OSNs) is significant for identifying the strong and weak ties. In this paper, we empirically investigate the evolution characteristics of Facebook and Wiki users’ social signature, capturing the distribution of frequency of interactions between different alters over time in ego network. The statistical results show that there are robust social signatures for collective egos based on interactions. Notably, individual social signature remains stable, no matter how alters change over time. Furthermore, we use structure information of OSNs to predict the social closeness in each interval in terms of embeddedness measurement, and find that alters identified from structural information still show the characteristic of social signature. The accuracy of the prediction in terms of embeddedness ranges from 33.47% to 66.90%. This work enables an effective method for identifying potential highly-related friends based on the regularity of social signature in online social network