Dynamic Social Networks: Search and Data Routing

Abstract This chapter focuses on two different dynamic processes that unfold based on connections existing in a social graph: (i) targeted social search and (ii) altruistic data routing. One key difference between these dynamic processes is the extent to which network information is exploited to determine next-hop intermediaries. For the targeted social search process with a local view of network connections, existing models are first reviewed and then some recent results of the authors providing a good statistical fit to empirical data are discussed. These results shed light on the prime utility of long-range and short-range connections in dynamically evolving targeted social search processes. In particular, the long-range connections are utilized to identify the target location up to a “small” region in the social search space quickly, whereas their navigational convenience disappears and the short-range connections become more helpful to approach the target for small social distances. This hierarchical interaction between two different types of social ties is the underlying root cause for the linear growth and saturation regimes observed in the delay function for targeted social search processes. In contrast to the targeted social search, the process of dynamic data routing in sociotechnological networks depends on the global understanding of existing connections (both physical ones at the device level and social ones at the individual level) between data sources and targets. Considering the global view of network connections, the process of data routing is formulated as an altruistic data routing problem taking social connections into account while allocating data rates over paths connecting sources and targets. The necessary and sufficient conditions for a strategy profile to be a group Nash equilibrium are provided for the process of altruistic data routing as well as showing the existence of at least one Nash equilibrium point in pure strategies.